The Regularization of Nonmarital Cohabitation: Rights and Responsibilities in the American Welfare State Grace Ganz Blumberg

Notre Dame Law Review Volume 76 Issue 5 Unmarried Partners and the Legacy of Marvin Article 2 v. Marvin

10-1-2001 The Regularization of Nonmarital Cohabitation: Rights and Responsibilities in the American Welfare State Grace Ganz Blumberg

Follow this and additional works at: http://scholarship.law.nd.edu/ndlr

Recommended Citation Grace G. Blumberg, The Regularization of Nonmarital Cohabitation: Rights and Responsibilities in the American Welfare State, 76 Notre Dame L. Rev. 1265 (2001). Available at: http://scholarship.law.nd.edu/ndlr/vol76/iss5/2

This Article is brought to you for free and open access by NDLScholarship. It has been accepted for inclusion in Notre Dame Law Review by an authorized administrator of NDLScholarship. For more information, please contact [email protected]. THE REGULARIZATION OF NONMARITAL COHABITATION: RIGHTS AND RESPONSIBILITIES IN THE AMERICAN WELFARE STATE

Grace Ganz Blumberg*

INTRoDucrIoN

Twenty years ago, I wrote a law review article criticizing the M1ar- vin' view that nonmarital cohabitation is most appropriately regulated by the rubric of contract.2 I suggested that stable, long-term cohabitation, like marriage, should give rise to a legal status entailing rights and responsibilities, both inter se and third-party, unless the cohabitants have made an enforceable agreement to the contrary.3 Since then, other American commentators have espoused this perspective, 4 and the foreign law sources upon which I drew for inspiration two decades ago have further developed their status-based treatment of

* Professor of Law, UCLA School of Law. I thank Donald R. Blumberg for his helpful editorial suggestions. 1 Marvin v. Marvin, 557 P.2d 106 (Cal. 1976). 2 See Grace Ganz Blumberg, Cohabitation Without Marriage: A Different Persprtihe, 28 UCLA L. REv. 1125, 1159-70 (1981). 3 See id. at 1167. 4 See, e.g., Cynthia Grant Bowman, A Feminist Proposal to Bring Back Common Law AMarriage,75 Op. L. Ruv. 709, 711-12 (1996); William A. ReppyJr., Property and Support Rights of Unmarried Cohabitants:A Proposalfor Creatinga New Legal Status; 44 L.-. L RE%, 1677, 1678 (1984); Amy Lir, Comment, In Defense of Washington's Equitable Treatment of PseudomaritalProperty, 29 IDAHo L. REv. 975, 995-98 (1992); Kathryn S. Vaughn, Comment, The Recent Changes to the Texas Informal MarriageStatute: Limitation or Aboli- tion of Common-Law Aariage?, 28 Hous. L Rv. 1131, 1147-49 (1991). For criticism of the application of contract analysis to marriage and marriage-like cohabitation, see, for example, Margaret F. Brinig, Status, Contract and Covenant, 79 CoPNT-L. L REV. 1573, 1594-99 (1994) (revieing MILTON C. REGa-N, JR., FAm.LY Lw,AND THE Puisurr oF Irns.Acy (1993)), and Ira Mark Ellman, The Theory of Alimony, 77 C . L REv. 1, 13-24 (1989). For an earlier critique of the unjust enrichment remedy, see Robert C. Casad, Unmarried Couples and Unjust Enridment"From Status to Contractand Bacb Again?, 77 Mica-. L. REx, 47, 49-51 (1978).

1265 1266 NOTRE DAME LAW REVIEW [VOL- 76:5 cohabitation.5 Yet, I can only modestly congratulate myself on my pre- science, for I did not even dimly foresee the path that law and society would take in regularizing nonmarital cohabitation in the United States. Although the developing substantive law of cohabitation is a matter of interest in itself, the story of how and, more speculatively, why we have moved toward that law is at least equally interesting. Le- gal developments are best understood contextually. In this Essay, I will attempt to link the particular character of the American welfare state with the developing American regularization of nonmarital cohabitation. A comparative look at convergent developments in our closest cultural relatives, English-speaking common-law countries, also sheds useful light on certain aspects of American family law and its relation to our version of the welfare state. The American welfare state (social security in the larger sense) consists of three interrelated institutions. The family is a highly effective system of wealth redistribution. In the family, income is generally redistributed from one adult to another (most often from men to women) and from adults to children. During the ongoing conjugal relationship, redistribution rarely requires any legal intervention. The economics of the household and, secondarily, ties of affection tend to assure ample redistribution. When a conjugal relationship dissolves by divorce, family law seeks to accomplish wealth redistribution by means of well-developed rules of property division, child support, and spousal support. When a marriage persists until the death of one of the parties, wealth redistribution is effected by the laws of intestacy, the common-law elective share, and community property distribution. Equally important is the American institution sometimes characterized as the "employee welfare state' 6 or "shadow welfare state."17 In-

5 See infra Part III.C. 6 See, e.g., David Charney, The Employee Welfare State in Transition, 74 TEx. L. REnv. 1601, 1601-02 (1996). 7 See, e.g., MARIE GoTrcuc, THE SHADOW WELFARE STATE: LABOR, BUSINESS, AND THE PoLITCS OF HEALTH CARE IN THE UNITED STATES 1-16 (2000). The "employee welfare state," or "shadow welfare state," overlaps but is not co. extensive with the "hidden welfare state" described by Christopher Howard in Ti-iE HIDDEN WELFARE STATE: TAX EXPENDITURES AND SOCIAL POLICY IN THE UNITED STATES (1997). Howard's definition of the "hidden welfare state" includes all tax expenditures with social welfare objectives, meaning those that parallel direct expenditures for income security, health care, employment and train- ing, housing, social services, education, and veterans' benefits. Familiar ex- amples include tax deductions for home mortgage interest and charitable contributions. Altogether, tax expenditures with social welfare objectives cost approximately $400 billion in 1995. Id. at 3. 2001] THE REGULARIZATION OF NONMARITAL COHABITATION 1267 cident to employment, employers provide workers and their families with a wide variety of essential welfare benefits, including health, disability, and retirement benefits. Thus, in addition to its role in the internal redistribution of cash income, the family acts as a conduit for benefits from the employee welfare state. Although the state is not involved as a direct provider in the employee welfare state, it under- writes employee benefit plans with generous tax subsidies conditioned upon employer maintenance of the essential social welfare chracter of the benefit plan, by adherence to nondiscrimination rules that de- ter employers from providing superior benefits to highly compensated employees. 8 Finally, and far less so than in most other Western countries, the state acts as a direct provider of social welfare. Direct provision is largely limited to categorical programs designed to operate, at most, as a safety net for certain classes of persons whose basic needs are not otherwise adequately met.9 This Essay traces the various and diverse strands of the developing law of cohabitation, with particular attention to their relationship to the American welfare state. The unusual character of the American welfare state tends to explain one noteworthy American development-employer extension of employment-related benefits to the nonmarital domestic partners of employees. The nature of the Ameri- can welfare state may also explain why the United States has been comparatively slow to recognize that cohabitants who claim and enjoy the benefits of marriage should also be expected to bear the burdens of marriage. The narrative of this Essay additionally uncovers two related themes. The first is the extent to which a relatively small sub- group of cohabitants-same-sex partners-spurred the movement to regularize nonmarital cohabitation. In the United States, same-sex couples are currently less than half as frequent as unmarried opposite- sex couples, 10 and the number of same-sex couples is relatively con- stant, while the number of unmarried opposite-sex couples is ever-

8 See I.RC. § 410(b) (1994). 9 The most prominent categorical assistance programs are Supplemental Secur- ity Income (SSI), 42 U.S.C. §§ 1381-1383(d) (1994), which provides benefits to ie indigent aged and disabled, and Temporary Assistance for Needy Families (TiNF), id. §§ 601-617 (Supp. IV 1999), which replaced Aid to Families With Dependent Chil- dren (AFDC), id. (1994). 10 In 1998, Census Bureau data show 55,303,000 married couples and 5,910,000 unmarried couples. Of the unmarried couples, 4,236,000 were opposite-sex couples, and 1,674,000 were same-sex couples. Stated othermise, there were eleven unmarried couples for every 100 married couples. Of the eleven unmarried couples, eight were opposite-sex couples, and three were same-sex couples. &e infra note 11, chart 2. 1268 NOTRE DAME LAW REVIEW [VOL- 76:5r increasing. 1 Nevertheless, in the United States, same-sex couples

11 The number of unmarried, opposite-sex couples rose eight-fold from 1970 to 1998. The Census Bureau did not begin to collect data on same-sex couples until 1990. The same-sex couple figures from 1994 to 1998 are relatively flat. Indeed, they show a slight decline in numbers (from 1,678,000 to 1,674,000). By contrast, the number of opposite-sex couples increased by sixteen percent during the same four- year period (from 3,661,000 to 4,236,000). In 1998, there were eight unmarried opposite-sex couples for every 100 married couples, up from one per 100 in 1970. More than one-third of these unmarried couples had a child under fifteen years of age in their household. The change is shown in the following table, taken from BUREAU OF THE CENSUS, CURRENT POPULA- TION REPORTS, SERIES P20-514, UNPUBLISHED TABLEs-MARITAL STATUS AND LIViNG AR. RANGEMENTS: MARCH 1998 (UPDATE) 71-73 tbl.8 (1998), at http://www.census.gov/ prod/99pubs/p20-514u.pdf (last visited August 30, 2001) (containing corrected and updated data through 1998 for information originally found in BUREAU OF THE CEN- SUS, CuRRENT POPULATION REPORTS, SERIES P-20-484, MARITAL STATUS AND LIVING AR- RANGEMENTS: MARCH 1994 (1996)).

TABLE A-8 UNMARRIED CouPLE HOUSEHOLDS, BY PRESENCE OF CHILDREN: 1970-1994 (1998) Unmarried Total Couples Per 100 Year Married Couples Unmarried Couples Married Couples Without With Children Children Total Under 15 Under 15 1998 55,303,000 4,236,000 2,716,000 1,520,000 8 1994 54,261,000 3,661,000 2,391,000 1,270,000 7 1990 53,256,000 2,856,000 1,966,000 891,000 5 1985 51,114,000 1,983,000 1,380,000 603,000 4 1980 49,714,000 1,589,000 1,159,000 143,000 3 1970 44,593,000 523,000 327,000 196,000 1

In 1990, the Census Bureau began collecting data on same-sex couples (as opposed to "roommates"). The data since 1994 for all couples show the following:

Opposite-Sex Total Married Couples Per Same-Sex Total Unmarried Couples Unmarried Couples 100 Married Couples Per 100 Couples Per 100 Year (1000s) (lO00s) Couples Married Couples Married Couples Opposite- Same- Sex Sex Total 1998 55,303 4,236 1,674 5,910 4 3 11 1996 54,667 3,958 1,684 5,642 7 3 10 1994 54,261 3,661 1,678 5,339 7 3 10

hi; BUREAU OF THE CENSUS, CURRENT POPULATION REPORTS, SERIES P20-496, UNPUn usrn TAB.Es-MARITAL STATUS AND LIVING ARRANGEMENTS: MARCH 1996 (UPDATE) 2001] THE REGULARIZATION OF NONMARITAL COHABITATION 1269 have been the dominant force in the movement to regularize nonmarital cohabitation. The second theme concerns the relationship between the private and the public sectors. In the United States, the government, particularly the federal government, often leads the way in claims to equal treatment and social justice, frequently by statutory regulation of a recalcitrant private sector.1 2 However, customary roles have been re- versed in the case of cohabitation. While the federal government leg- islated "in defense of marriage,"13 private sector employers were extending family benefits, previously limited to lawfully married spouses of employees, to same-sex partners and sometimes to nonmarital opposite-sex partners as well.14 Similarly, family benefits granted to an employee's step-children (that is, the children of the employee's spouse) were also frequently extended to the children of an employee's same-sex or opposite-sex partner.15 These two secondary themes are largely local; they are particular aspects of the American regularization of cohabitation. The larger theme is comparative in scope. It examines the connection between the law regulating the obligations of family members to one another

71-73 tbl.8 (1998), at http://vwv.census.gov/prod/3/98pubs/p2-496u.pdf (last 'is- ited August 30, 2001); BUREAU OF THE CENSUS, CuRRENT POPUL,%TION RErorrs, SEIES P20-484, AR.rruAL STATUS AND LIVING ARRANGEIENTS: AARCH 1994 at 71-73 tbl.8 (1996), availableat http://wv.census.gov/prod//pop/p2484.pdf (last visited Au- gust 30, 2001). Same-sex couples are almost half as frequent as unmarried opposite- sex couples. At any given time, nearly half of gay men and more than half of lesbians report themselves in a relationship with a primary partner, and many of these have exchanged rings or othenvise engaged in a commitment ceremony. See sources cited in David L Chambers, What I? The Legal Consequencesof Mlarriageand the Legal Needs of Lesbian and Gay Male Couples, 95 MicH. L RE%. 447, 449 nn.3-6 (1996). 12 See, e.g., Education Amendments of 1972, tit. 9, 20 U.S.C. § 1681 (1994) (prohibiting sex discrimination in any education program or activity receiving federal financial assistance); Equal Pay Act of 1963, 29 U.S.C. § 206 (1994) (prohibiting sex- based wage discrimination); Civil Rights Act of 1964, tit. 7, 42 U.S.C. § 2000e to e-17 (1994) (prohibiting discrimination in employment on the basis or race, color, religion, sex, or national origin). 13 Defense of Marriage Act (DOMA), Pub. L No. 104-199, 110 Stat. 2419 (1996) (codified as amended at 28 U.S.C. § 1738C (Supp. IV 1998) (providing that states are not required to recognize sister state same-sex marriages) and 1 U.S.C. § 37 (Supp. IV 1999) (providing that the federal government does not recognize same-sex marriages)). The enactment of DOMA was prompted by Baehr v. Leuin, 852 P.2d 44 (Haw. 1993), and Baehr v. Miike, No. CIV.91-1394, 1996 WL 694235 (Ham Cir. Ct. Dec. 3, 1996), affd, 950 P.2d 1234 (Haw. 1997). Baehrv. Lxewin is discussed infra text accompanying notes 36-43. 14 See infra notes 75-78 and accompanying text. 15 The University of California plan, described iifta at notes 82-101 and accompanying text, is illustrative. See infra note 92. 1270 NOTRE DAME LAW REVIEW [VOL. 76-,r

and the state's historical and cultural role as a primary provider or secondary guarantor of social welfare. It explores the relationship between growing pressures on the welfare state (largely resulting in the decline of the welfare state) and the development of a facially robust private law of rights and responsibilities at divorce and at the termination of opposite-sex and same-sex nonmarital cohabitation. This relationship is evidenced in countries that are culturally closest to the United States, such as Canada, New Zealand, Australia, and England. In these countries, the trend has been to privatize economic obligations, that is, to substitute property division and family support obligations for public social security.16 The substitutionary role of private law ultimately paves the way for uniform inclusion of all cohabiting couples, whether formally married or not.17 Unlike most other developed countries, the United States has never committed itself to the comprehensive goals of a fully developed welfare state. Consequently, it is not ordinarily thought to be the role of government to guarantee the social welfare of the citizenry.' 8 This perspective may affect the way that the United States has conceptualized and rationalized family

16 For accounts of changing child support policy in England, seeJane Lewis, Fain- ily Policy in the Post-warPeriod, in CROSS CuRRENTrs: FAMILY LAW AND POLICY IN THE US AND ENGLAND 81, 94-100 (Sanford N. Katz et al. eds., 2000), and Mavis Maclean, Dce- galized Family Obligations, in FAMILY LAW AND FAMILY PoucY IN THnE NEw EUROPE 129, 130-36 (Jacek Kurczewski & Mavis Maclean eds., 1997). For a contrasting account of American law, see Barry L. Friedman & Martin Rein, The Evolution of Family Policy in the United States after World War II, in CRoss CuRRENTs: FAMILY LAW AND POLICY IN THE US AND ENGLAND, supra, at 101, 110-11. 17 Writing about property division and spousal support in England, John Eeke- laar observes: As far as the law on financial provision is concerned, the disjuncture between the formal basis of the law (dissolution of marriage) and its underlying substance (sharing of resources acquired during a failed common enterprise-specifically the task of beginning to bring up a child) seems to be becoming more visible. The extension of the substance in due course to unmarried cohabitants who also have children seems irresistible. John Eckelaar, Post-divorce FinancialObligations, in CROSS CuRRENTrs: FANILY LAW AND POLICY IN THE US AND ENGLAND, supra note 16, at 405, 421. Although England has not systematically applied its marriage laws to cohabitants, Gillian Douglas observes: Two basic approaches can be discerned. The first was to give cohabitation some recognition as a distinct status, and to equate it with marriage by giving to cohabitants the same legal benefits and burdens that are given to spouses. Gillian Douglas, Marriage, Cohabitation, and Parenthood-From Contract to Status, in CROSS CuRRENTs: FAMILY LAW AND POLICY IN THE US AND ENGLAND, supra note 16, at 211, 213. Those benefits and burdens include entitlement to succeed to a tenancy, to claim provision from a decedent's estate, and to wrongful death benefits. Id. 18 See Friedman & Rein, supra note 16, at 111-21. 2001] THE REGULARIZATION OF NONMARITAL COHABITATION 1271 law obligations, as compared to countries that have experienced the content and ethos of a more fully realized public welfare state.19 Spe- cifically, differing perspectives may explain significant differences between the United States and our cultural cousins in the legal treatment of nonmarital cohabitation. I approach this subject as a legal scholar who, in the course of professional life, has also participated in two of the legal developments discussed in this Essay- the extension of health benefits to employees' same-sex domestic partners and the drafting of the American Law Institute's Principlesof the Law ofFamily Dissolution chapter on the inter se obligations of nonmarital cohabitants. I am not directly af& fected by any of these developments. Instead, I am a long-married spouse and parent who has had the better part of a lifetime to reflect on the family and its myriad legal incidents and, consequently, to feel concern for others who live in family circumstances indistinguishable from those of my family but, through happenstance or legal disability, may nevertheless be denied the protective cover of the lawfully consti- tuted family. Additionally, I am a long-time participant in the govern- ance of an elaborate employee welfare scheme and, together with my family, a beneficiary of that scheme. In both capacities, I have frequent occasion to feel concern for those who are similarly situated, but for various reasons do not enjoy the protective cover of my particular employee welfare state.

I. SORTING THE VARIOUS STRANDS OF THE DEVELOPING LVW OF COHABrrATION The law and social treatment of cohabitation has been developing in a number of diverse ways, so that it becomes increasingly difficult to broach any aspect of the subject without first carefully specifying and cabining the precise topic of discussion. Ultimately, there may be an underlying coherence and a unifying set of issues, but the possibility of confusion is sufficiently great to require an initial sorting of those different developments. The diverse strands of the legal regulation of nonmarital cohabitation can be organized in various ways. For purposes of the themes of this Essay, I employ the organizational rubric of "rights and responsibilities." Of course, in any reciprocal relationship, one party's right may be the other's responsibility. Nevertheless, aspects of the move-

19 See Guy STANDING, GLOBAL LABOUR FLE-XIBILrrr SEEING DLnBtuTmnJusTcE 251-91 (1999); LouAs TsouKAus, THE NEv EuROPEAN EcoNo. w RE'isrVIE 114-37 (1997). See generally EuROPEAN SOctAL PoLicy. BEnwr.-N FR GME nWn ANTD INTEur,. TION (Stephan Liebfried & Paul Pierson eds., 1995). 1272 NOTRE DAME LAW REVIEW [VOL. 76:5 ment to regularize cohabitation may helpfully be understood as initiatives either to acquire rights or, contradistinctively, to assign responsibilities. Not surprisingly, it is usually individuals who seek to establish fights and the state that seeks to assign responsibilities. In marriage, rights and benefits accrue to the parties during their relationship by virtue of their marital status. Although the parties nominally have legal obligations to each other during an ongoing marriage, those obligations largely go unnoticed because they are consistent with ordinary spousal behavior in the course of the parties' life together. At dissolution of the relationship, particularly at dissolution by divorce, those obligations take on significance in the allocation of property and the duty to pay child support and spousal support. Al- though there is ordinarily a temporal disjunction between a spouse's experience of the rights and obligations of lawful marriage, marriage may be understood as a package, variously conferring benefits and burdens upon persons who make a life together. The first and most widely reported strand is the creation of a new legal status that falls short of marriage, but that entails some of the public rights and benefits of marriage. This status is often called "domestic partnership." The intended beneficiaries of this status are largely same-sex couples who are not eligible to marry, although in some instances the opportunity has also been made available to opposite-sex couples. This development is embodied in state and municipal domestic partnership statutes, which define the incidents of domestic partnership and allow couples to register as domestic partners. Denmark enacted comprehensive domestic partnership legislation in 1989, and similar legislation was subsequently adopted by many other western European countries. In the United States, domestic partnership legislation has been quite modest, except in Vermont and Hawaii, where broad-sweeping legislation has been required or impelled by successful equal protection litigation. The second strand, which has been less commented upon than any other but has considerable strength in the United States, is the extension to nonmarital cohabitants of privaterights and benefits historically enjoyed only by married persons. This trend has been most noteworthy in the employer-employee relationship, where many employers have extended family benefits to the domestic partners of their employees. For this purpose, employers generally require a declaration that the parties live together and assume some sort of mutual responsibility for each other. Extension of employment benefits to domestic partners is more meaningful in the United States than it would be in some other Western countries, because in the United States much basic social security is provided as an aspect of employ- 2001] THE REGULARIZATION OF NONMARITAL COHABITATION 1273

ment compensation. Employment-based social security often provides not only for the employee alone, but also for family members with whom the employee shares a household.20 This is most notably and distinctively true of health care and old-age pensions. The third strand, and the one that has the longest legal history, both here and abroad, is the legal obligation that one nonmarital partner may have to the other at the dissolution of their relationship by death or inter vivos separation. This is the subject of the contract rubric expressed by Marvin and its progeny.21 Otherjurisdictions and authorities have taken the path of status rather than contract; that is, they treat persons who live together in stable marriage-like cohabitation as though they are married, despite the absence of the legal for- malities of marriage. The latter approach is expressed in the case law of Washington State, 22 the legislation of other countries,2 and the American Law Institute's Principles of the Law of Family Dissohtion.2 4 The first two strands can be understood as claims to equal access to rights and benefits generally accorded to married persons. In re- questing the extension of benefits to nonmarital cohabitants, an employee who resides with a nonmarital cohabitant is making a claim to equal treatment with a co-worker who resides with a lawful spouse. Similarly, the domestic partnership movement is intended to secure as many of the incidents of marriage as the state is willing, or constitu- tionally compelled, 25 to allow. Each development requires that claim- ants affirmatively assert and acknowledge their relationship as nonmarital cohabitants. The stronger form of affirmation is domestic partnership legislation, which requires that persons register formally as domestic partners. Employers generally require some weaker form of affirmation. Parties registering as domestic partners seek to identify themselves as such and to obtain rights and benefits that accompany that status. Although some domestic partnership statutes may also impose inter se and third-party (creditor) obligations,2 6 the enrolling parties

20 It sometimes even provides for family members with whom the worker does not share a household, such as children from a prior marriage who reside ith the other parent. 21 Marvin v. Marvin, 557 P.2d 106 (Cal. 1976). 22 See infra note 105. 23 See infra notes 138-56 and accompanying text. 24 PRINCIPLES OF THE LAW OF FAMILY DissoLrrION: ANALsYs AND REcoMEND.A- TiONS (Tentative Draft No. 4, 2000) [hereinafter ALI PINCIPLES (Tentative Draft 2000)]. 25 See infra Part MA 26 See, ag., infra text accompanying note 65. 1274 NOTRE DAME LAW REVIEW [VOL- 76:0t are likely to see these obligations as incidental, remote, and contingent, much as persons entering marriage generally view the possibility of post-dissolution obligations. This is even more the case with employee benefits, where the employee claiming benefits for a nonmarital partner is generally required only to attest to the existence of the relationship, that is, to the fact of sharing a household with a 2 7 nonmarital partner in some measure of economic interdependence. Employee benefits represent, most clearly, a right or benefit without any corresponding obligation running from one cohabitant to the 8 other.2 Unlike the first two developments, the third strand, traditionally operative at the dissolution of the relationship by death or separation, entails inter se claims, that is, claims by one party against the other. One party is asserting a right and, if the right is recognized, the other party may experience a corresponding responsibility or burden. I shall therefore refer to this third area as the "responsibilities" strand in that it has the effect of imposing responsibilities upon persons who might not otherwise voluntarily undertake them. This is the strand addressed by Marvin and by alternative status-based treatments. It does not require or depend on any affirmative registration as domestic partners, but instead looks to the parties' inter se behavior, either contractual in the Marvin rubric or social in the status rubric. Although these three strands are distinct, they are not unrelated. For example, domestic partnership registration and declaration may be treated as some evidence of the parties' social relationship, for the purpose of imposing obligations, either contractual or status-based, under the third strand of cohabitation doctrine.2 9 More importantly,

27 See, for example, the UCLA affidavit for same-sex cohabitants, reprinted infra at text accompanying note 94. 28 The extension of employment-based benefits to the nonmarital partner of an employee is not based on the rationale that the employee has any legal duty to support the nonmarital cohabitant, for the employee ordinarily does not. (By contrast, this is a frequent rationale for the extension of public and private benefits to family members of an employee. From this perspective, the equality claim is unavailing because the married employee and the cohabiting employee are not equally situated. One owes a duty of support to his spouse; the other generally does not owe a duty of support to his cohabitant.) Instead, the extension of benefits is more plausibly based on the economics of a family household. Whether or not cohabitants who share a household are legally bound to do so, they can be expected to provide each other with support. This rationale also justifies the occasional extension of health benefits to a dependent aged parent or adult child who resides with a covered employee. The core notion is that any householder can reasonably be expected to support family members with whom he or she shares the household. 29 See, e.g., ALI PmINcIpus (Tentative Draft 2000), supra note 24; § 6.03(7) 0). 2001] THE REGULARIZATION OF NONMARITAL COHABITATION 1275 the coexistence of these three distinct strands raises issues about the relationship between rights and responsibilities in the legal and social regulation of cohabitation.

II. EQUALrIY CLAMS TO THE RIGHTS AND BENEFITs OF MARRIAGE

A. Equality Claims to Lawful Marriageand Domestic Partnership The equality claim to the rights and benefits of laufuI marriage began boldly, although unsuccessfully, in the early 1970s. In a series of cases, state courts uniformly held that the restriction of marriage to opppsite-sex partners posed no credible equal protection issues. 0 In the two decades that followed, the organized gay and lesbian community engaged in a spirited debate on whether it ought to seek access to the status or incidents of marriage. Some took the view that gay persons generally should not replicate constricting heterosexual institutions and behavioral patterns, such as marriage and lifetime commitment to another person, but should instead be free spirits in an otherwise overly constrained society.3' Another view, agnostic or even positive about long-term commitment, nevertheless disdained marriage as an institution based on and embodying patriarchal privi- lege, that is, as an institution that systematically oppresses women.32 (Not surprisingly, this perspective was more frequently espoused by lesbians than by gay men.3 3 ) On the other hand, many same-sex couples desired the regularization of their relationships as well as the

30 Se e.g.,Jones v. Hallahan, 501 S.W.2d 588, 590 (Ky. 1973) (rejecting the claim of same-sex marriage applicants who invoked the constitutional rights of free exercise of religion, freedom of association, and freedom to marry); Baker v. Nelson, 291 N.W.2d 185, 186-87 (Minn. 1971) (rejecting same-sex marriage applicants' equal protection and due process claims); Singer v. Hats, 522 P.2d 1187, 1195-97 (Wash. Cit. App. 1974) (rejecting the claim of same-sex marriage applicants despite an equal rights amendment to the state constitution). For contemporaneous commentary, see Comment, The Legaliy of Homosexual Marriage,82 Yu. LJ. 573, 574-83 (1972), and Note, Homosexuals'Rightto Maly: A ConstitutionalTest and a LegislativeSolution, 128 U. PA. L REV. 193, 197-213 (1979). 31 See generally, ag., Paula L. Ettelbrick, Since Illzen Is Marriagea Path to Liberation?, OUT/LOOK, Fall 1989, at 9. 32 See, eg., Nancy D. Polikoff, Ie WIMl Get 117at IWe Ask Bor I7ay Legalizing Gqrj and Lesbian Marriage Wdl Not "Dismantlethe Legal Structure of Gender in Ever Moariage," 79 VA. L. REv. 1535, 1536 (1993). 33 SeeWxu. i N. ESKRmDE, JR., THE CASE FOR S.%ME-sEX M, ,RRIO.:FROM SExU ,u LmERwvTo Crvusna_ Coimrrnr'r 75-77 (1996); Thomas B. Stoddard, 179y Gay People Should Seek the Right to Many, OUT/LOOK, Fall 1989, at 9, 10-13. But see Steven K. Homer, Note, Against Marriage, 29 HARv C.R.-C.L L RE% 505, 530 (1994). See generally ADRm.v SuuVAN, VIRTUALLY Noatu.: AN ARUcMErr ABoUT Ho.tosEw-un,rv (1995). 1276 NOTRE DAME LAW REVIEWV [VOL-. 76:5 various rights and benefits that accrue to married persons from marital status. 34 Ultimately, the latter view prevailed. Lambda Legal De- fense and Education Fund, the legal arm of the organized gay and lesbian movement, pursued litigation in Hawaii, an historically toler- ant state with a liberal supreme court. The strategy was to secure lawful same-sex marriage in one or more hospitable states. Same-sex residents of other states could then lawfully contract marriages in one of those hospitable states and, arguably, rely on ordinary choice-of-law principles to secure legal recognition of their same-sex marriage in their home states.35

34 See sources cited supra notes 31-33. On the face of it, the pro argument was more powerful than the con argument in the sense that mere guarantee of access does not require participation in a legal institution. Even with access to the status or incidents of marriage, gays and lesbians could still choose to decline both as well as long-term monogamous relationships. However, the movement to regularize same- sex cohabitation, once undertaken, is likely to reach the relationships of members of the anti-regulation group as well. See infra Part III. 35 According to the Restatement, A marriage which satisfies the requirements of the state where the marriage was contracted will everywhere be recognized as valid unless it violates the strong public policy of another state which had the most significant relationship to the spouses and the marriage at the time of the marriage. REsTATEMENT (SEcOND) OF CoNrar OF LAws § 283(2) (1971). The strong public policy provision generally comes into play when the marriage would violate some criminal law of the state. In most American states, same-sex marriage would not violate any criminal law. This choice-of-law principle would ordinarily apply even though the home state would not itself allow the parties to contract a same-sex marriage in the home state. After the early same-sex marriage cases of the 1970s, a number of states redrafted their marriage statutes to specify that a marriage may only be contracted between a man and a woman. See, for example, CAL. FAM. CODE § 306 (West 1994), which de-' fines marriage as "a personal relation arising out of a civil contract between a man and a woman." In 1977, the California legislature added the words "between a man and a woman" to signify that same-sex couples cannot be lawfully married in Califor- nia. See 1977 Cal. Stat. 339. Nevertheless, California courts ordinarily recognize a marriage lawfully contracted in anotherjurisdiction, even though the marriage would not satisfy California marriage requirements if it were contracted in California. Call- fornia courts regularly recognize, for example, common-law marriages validly contracted elsewhere, even though California abolished common-law marriage in 1895 and therefore would not recognize a purported common-law marriage contracted in California after that date. See, e.g., In re Marriage of Smyklo, 226 Cal. Rptr. 174 (Ct. App. 1986) (recognizing an Alabama common-law marriage). To obviate the possibility that California courts might recognize same-sex marriages contracted elsewhere, the California electorate approved Proposition 22 in March 2000. SeeJenifer Warren, Proposition 22: Ban on Gay MarriagesWins in All Regions but Bay Area, L.A. TMEs, Mar. 8, 2000, at A23. 2001] THE REGULARIZATION OF NONMARITAL COHABITATION 1277

Invoking the Equal Protection Clause of the Hawaii Constitution, in Baehr v. Lewin,36 the Hawaii Supreme Court held that the state must satisfy the very highest standard of constitutional review, "the strict scrutiny standard," in order to justify restricting marriage to opposite- sex couples.37 Entirely avoiding discussion of sexual orientation, the Hawaii Supreme Court characterized the issue as one of simple sex discrimination: persons of one sex are denied the right to marry persons of that sex, a right accorded to persons of the other sex.3 On remand to the trial court, the State failed to meet the heavy constitutional"burden; thus, the State of Hawaii was required to make marriage equally available to gay and lesbian couples.3 9 However, the voters of Hawaii subsequently authorized their legislature to amend the state constitution to overrule Baehrinsofaras it required Hawaii to recognize same-sex marriage. 40 As an apparent quid pro quo, the Ha- waii legislature provided "reciprocal beneficiary" legislation for persons legally unable to marry. In a move as oblique as the sex discrimination rationale of Baehr, the Hawaii legislature side-stepped the issue of same-sex relationships by extending relief to any two unmarried individuals legally prohibited from marrying each other under state law.41 The legislative findings refer to brothers and sisters, a widowed mother and her unmarried son, and, finally, "two individuals who are of the same gender."42 The effect of the Hawaii legislation is to extend to same-sex couples some of the third-party benefits available to married couples, such as eligibility for derivative health insurance coverage for family members, hospital visitation privileges, wrongful death and loss of consortium claims, family and funeral leave from employment, and state employee pension rights and death

36 852 P.2d 44 (Haw. 1993). 37 Id. at 67. 38 Id. at 64. The court found inspiration in Lotting v. Virginia, 388 U.S. 1, 12 (1967), which may be understood to give a similar reading to the Equal Protection Clause of the Fourteenth Amendment. Striking dowm Virginia's anti.miscegenation laws on the grounds of equal protection and substantive due process, the United States Supreme Court rejected Virginia's argument that its prohibition of interracial marriages satisfied the Equal Protection Clause because it treated all races equally in that no person of one race was allowed to marry a person of another race. Id. at 8. 39 Baehr v. Miike, No. CIV.91-1394, 1996 WL 694235, at *21 (Haw. Cir. Ct. Dec. 3, 1996), afJ'd, 950 P.2d 1234 (Haw. Dec. 3, 1997). 40 SeefDeb Price, Eekdions Rect RepublicanExtremism, DrR-orr NmEvs, No. 9, 1998, at All. 41 HAW. REv. STAT. ANN. §§ 572C-1 to -7 (Michie 1999). 42 1997 Haw. Sess. Laws, Act 383, (H.B. 118) § 2. Even so, the legislation appears to have been controversial. It became law without the Governor's signature. &relHmw. REv. STAT. ANN. § 572C-2. 1278 NOTRE DAME LAW REVIEW [VOL- 76:5 benefits. The legislation includes reciprocal beneficiaries as spouses for state heirship and elective share legislation and the pretermitted spouse statute.43 Despite recognizing claims of one reciprocal beneficiary against another when their relationship ends at death, the Ha- waii legislation creates no claims at inter vivos termination of the relationship. This gap is understandable in view of the broad scope of the legislation. One would not want to impose continuing obligations, for example, upon an unmarried son who, in order to marry, terminates a reciprocal beneficiary relationship with his mother. However, the nature of the relationship between persons who cannot marry because of consanguinity and the relationship of persons who cannot marry because they are of the same sex is, or at least should be, distinctly different. An unmarried person who has access to benefits may commendably wish to share them with a less fortunate blood relative. That unmarried person should be able to terminate the relationship to the less fortunate relative without continuing obligation. On the other hand, same-sex partners who register as reciprocal beneficiaries are much more likely to be indistinguishable from opposite-sex couples who marry. Their relationship may include children as well as a much greater degree of financial interdependence than is likely in the case of reciprocal beneficiaries who are blood relatives. Thus, Ha- waii's oblique treatment is one that is not fully satisfactory to persons other jurisdictions would characterize as "domestic partners." The campaign for gay and lesbian marriage next looked to Ver- mont. The Vermont Constitution includes a "common benefit" clause, which provides that the government is "instituted for the common benefit, protection, and security of the people, nation, or community, and not for the particular emolument or advantage of any single person, family, or set of persons, who are a part only of that community."44 In Baker v. State,45 the Vermont Supreme Court held that the common benefit clause was violated by Vermont's exclusion of same-sex couples from the benefits and protections that Vermont provides to opposite-sex married couples. 46 The court concluded that the Vermont legislature had two options: it could extend the status of marriage to same-sex couples; or it could extend to same-sex couples the benefits and protections provided to married couples. 47 The legis-

43 See HAv. REV. STAT. ANN. § 572C-3 to -7. 44 VT. CONST. ch. 1, art. 7. 45 744 A.2d 864 (Vt. 1999). 46 Id. at 886. 47 Id. 2001J THE REGULARIZATION OF NONMARITAL COHABITATION 1279 lature chose the second option, enacting comprehensive domestic partnership legislation for same-sex couples. The Vermont legislation creates legal parity, to the extent state law alone can do so, between married couples and same-sex couples who enter a legally regulated "civil union."48 The legislation extends all the state law rights and obligations of marriage to a civil union of same-sex partners.4 9 The substantive and procedural requirements of entry are the same for marriages and civil unions.50 Relationships may not be incestuous5l or bigamous and must be duly licensed, ceremoni- alized, and registered.5 2 The dissolution of either is adjudicated in family court, where marital rights and obligations apply equally to marriages and civil unions.53 Parties to a civil union are treated equally with spouses with respect to death rights, wrongful death ac- tions, adoption law and procedure, group insurance for state employees, spousal abuse programs, workers' compensation benefits, family leave benefits, state and municipal taxes, the laws determining the parenthood of children born during the relationship of the parties, and a host of lesser incidents of marriage.54 The legislation also regulates insurers. Any insurance policy that provides coverage to married couples, spouses, and families must provide equivalent coverage for parties to a civil union and their families.55 The Vermont legislation recognizes that parties to a Vermont civil union will not be treated equally with spouses under federal income tax and estate tax law.50 However, for Vermont income tax purposes, the legislation treats parties to a civil union as though federal income tax law recognizes a civil union in the same manner as Vermont law.5 7 The legislation also con- forms the Vermont estate tax to federal law.58 From the point of view of family law, the distinction between a full-blown civil partnership, such as Vermont's civil union, and a law-

48 An Act Relating to Civil Unions, 1999 Vt. Acts & Resolves 91 (codified in scattered sections of VT. ST. ANN.). 49 VT. STAT. ANN. tit. 15, § 1204 (2000). 50 Id. § 1202; VT. STAT. ANN. tit. 18, §§ 5160-5169 (2000). 51 VT. STAT. ANN. tit. 15, § 1203. Compare the Havaii legislation, which requires ineligibility to marry on account of, inter alia, consanguinity. The Haiwii legislation may be read, darkly, to foster domestic relationships that, if they become sexual, would violate Hawaii's incest laws. HAw. REV. STAT. ANN. § 5720-4 (Michie 1999). 52 VT. STAT. ANN. tit. 15, § 1202; VT. STAT. ANN. tit. 18, §§ 5160-5169. 53 VT. STAT. ANN. tit. 15, § 1206. 54 Id. § 1204. 55 VT. STAT. ANN. tit. 8, § 4063a (2000). 56 See VT. STAT. ANN. tit. 32, § 7401(a) (2000). 57 Id. §§ 3001, 5812. 58 See id §§ 3001, 5812, 7401(a). 128o NOTRE DAME LAW REVIEW (VOL- 76:5 ful marriage is merely symbolic. In law, marriage is simply the sum of its legal incidents. However, from the perspective of religion and social psychology, there are other dimensions to marriage. The maintenance of marriage as an exclusively heterosexual opportunity appears to have important symbolic meaning for persons who are nevertheless amenable to the notion of domestic partnership for same-sex couples.59 Although no other American legislation has matched the scope of the Vermont civil union statute, one state60 and many municipali- ties 1 have enacted domestic partnership legislation. California, for example, permits state registration of "domestAic partnerships" com- posed of two unmarried adults who are either of the same sex or, if they are of the opposite sex, are over the age of sixty-two and qualify for social security benefits. 62 Partners must share a common residence, agree to be jointly responsible for basic living expenses in- curred by either of them during their relationship, and file a declaration of domestic relationship with the Secretary of State. 3 The legislation includes provisions for registering and terminating domestic partnerships. 64 Registration creates potential liability to creditors and a few basic rights. 65 Health facilities must treat a domestic partner and the children of a domestic partner as family members for purposes of hospital visitation. 66 The legislation also amends the Gov- ernment Code to authorize state and local government employers to offer health care coverage and related benefits to the domestic partners of state and local government employees. 67 This legislation may

59 California, for example, enacted relatively modest domestic partnership legislation before the voters approved Proposition 22, which provides that California will not recognize same-sex marriages lawfully contracted elsewhere. See 1999 Cal. Stat. 588. The proponents of Proposition 22 emphasized that they did not oppose domestic partnership legislation; their goal was merely to preserve the special status of marriage for opposite-sex couples alone. See David 0. Coolidge, MarriageIs Not Meant for Same-Sex Couples, L.A. TIMEs, Feb. 28, 2000, at B5. 60 See text accompanying infra notes 61-68. 61 See Craig A. Bowman & Blake M. Cornish, Note, A More Perfect Union: A Legal and SocialAnalysis of Domestic PartnershipOrdinances, 92 COLUM. L. Rv. 1164, 1188-92 (1992) (discussing the municipal ordinances). 62 1999 Cal. Stat. 588. 63 Id. 64 See id 65 See id 66 See id 67 See id. (adding CAL. FAM. CODE § 297 (West Supp. 2001), CAL. GOV'T CODE § 22867 (West Supp. 2001), and CAL. HEALTH & SAFETY CODE § 1261 (West 2000)). 2001] THE REGULARIZATION OF NONMARITAL COHABITATION 1281 be understood as California's first step in extending some of the rights and obligations of marriage to domestic partners.8 The gay and lesbian initiative for same-sex marriage has not yet accomplished its ambitious constitutional and choice-of-law project. However, law and society have not responded with offhand rejection, as they did to the constitutional litigation of the early seventies.6 9 In- stead, they effectively made a counter-proposal of domestic partnership. This process was evidenced most dramatically in the test state of Vermont. In Vermont, the state supreme court gave the legislature a choice between marriage and substantially equivalent domestic partnership,70 and the legislature opted for the latter.71 Jurisdictions that were not challenged by successful constitutional litigation have tended to adopt weaker forms of domestic partnership legislation, often relatively ad hoc in nature.72 In contrast, strong domestic partnership legislation has been extensive in Western Europe. Some countries have introduced registered partnership for same-sex couples only (Denmark, Nonvay, Sweden, and Iceland). Others have introduced registered partnership for all couples (Netherlands, France, and the Spanish provinces of Catalunya and Aragon). In most European countries, partnership registration has almost all of the consequences of marriage, except for

68 See, for example, A.B. 25, 2001-02 Leg., Reg. Sess. (Cal. 2001), which would extend the following rights historically reserved for spouses to registered domestic partners: derivative tort claims, including loss of consortium and %%rongful death claims; right to make medical treatment decisions for each oher;, exension of derivative group health insurance to the domestic partners of subscribers on the same terms normally provided for a subscriber's dependents; extension of sick leave, when provided to an employee to care for a spouse or child, to the provision of care for an employee's domestic partner or child of a domestic partner;, extension of spousal con- servatorship rules to domestic partners; extension of statutory will, intestacy rights, and administration rights to domestic partners; exclusion from taxable income, under California income tax law, of employer-provided health insurance for a domestic partner of an employee (an exclusion not allowed by federal income tax law, see infra note 87); and, for purposes of eligibility for unemployment compensation benefits, "good cause" for leaving employment includes accompanying one's domestic partner to a new location. Assembly Bill 25 is the "modest" bill. The more expansive A.B. 1338, 2001-02 Leg., Reg. Sess. (Cal. 2001), would follow the Vermont model, se text accompanying supra notes 48-58, and create a civil union status that provides all the rights and obligations of marriage. 69 See supra note 30 and accompanying text. 70 Se Baker v. State, 744 A.2d 864, 867 (Vt. 1999). 71 See An Act Relating to Civil Unions, 1999 Vt. Acts & Resolves 91 (codified in scattered sections of VT. STAT. ANN.). 72 See supra notes 60-68 and accompanying text. 1282 NOTRE DAME LAW REVIEW [VOL- 76:5 rights pertaining to the adoption of children.73 These countries and jurisdictions are, in historical order, Denmark (Danish Registered Partnership Act, 1989); Norway (Norwegian Registered Partnership Act of 1993); Sweden (1995); Iceland (1996); Greenland (1996); Netherlands (1998); Catalunya, Spain (1998); Aragon, Spain (1999); and France (PACS 1999).74

B. Equality Claims to Employment Benefits Independently of any legislation or court action, and for that reason generally less observed, American employers have increasingly been treating nonmarital cohabitants equally with married persons for purposes of employee benefits.7 The trend is particularly pro- nounced with large corporations, 76 educational institutions, 77 and

73 From the American perspective, the European exclusion of lights pertaining to the adoption of children is noteworthy. Even before the introduction of any notion of domestic partnership, American law was relatively permissive on gay and lesbian adoption of children, both with respect to children who were fully adoptable and the children of a same-sex domestic partner. See sources cited in ALl PRINCIPLES (Tentative Draft 2000), supra note 24, § 6.03, reporter's n. cmt. d. 74 The European statutes are collected in REcoGNISING SAME-SEx RELATIONSHIPS 11-16 (Law Commission of New Zealand, Study Paper 4, 1999). 75 In 1982, the Village Voice, a New York City weekly newspaper, became the first employer to offer domestic partner benefits to gay and lesbian employees. In 1992, Lotus Development Corporation became the first publicly traded company to offer domestic partner benefits. A 1997 survey by KPMG Peat Marwick found that thirteen percent of U.S. employers extended health care benefits to domestic partners. By 2001, more than 2500 public and private employers had extended health care benefits to domestic partners. See Human Rights Campaign, Wat Are Domestic PartnerBene- fits, at http://www.hrc.org/worknet/dp/dp-whatisdp.asp (last visited Apr. 7, 2001). 76 See id. The Human Rights Campaign reports that 116 of the Fortune 500 companies offer domestic partner health benefits, including six of the top ten companies, Human Rights Campaign, Fortune 500 Companies That Offer Domestic PartnerHealth Ben- efits, at http://vvw.hrc.org/equalityatexxon/fortune.asp (last visited Feb. 22, 2001). Data on whether the benefits are extended only to same-sex partners or also to opposite-sex domestic partners are incomplete. In companies for which such data are available, approximately forty percent offer benefits to same-sex partners only; the remaining sixty percent offer benefits to same-sex and opposite-sex domestic partners. Id. 77 The Human Rights Campaign reports that 129 colleges and universities offer domestic partner health benefits. Human Rights Campaign, Colleges and Universities That Offer Domestic PartnerHealth Benefits, at http://www.hrc.org/worknet/index.asp (last visited Feb. 22, 2001). Data on whether the benefits are extended only to same- sex partners or also to opposite-sex domestic partners are incomplete. For colleges and universities for which such data are available, approximately fifty-eight percent offer benefits to same-sex partners only; the remaining forty-two percent offer benefits to same-sex and opposite-sex domestic partners. Id. 2001J THE REGULARIZATION OF NONMARITAL COHABITATION 1283 state and municipal employers.78 The significance of this development lies in the central role that employment benefits play in the American welfare system. By welfare system, I mean not merely "welfare" in the popular sense (public assistance), but rather the entire structure of rights and obligations that secure the welfare of individuals and families in the United States. After World War II, while other countries were establishing national health systems, American employers increasingly provided health insurance as an employment benefit for workers and their families. 79 Although this approach has been supplemented by relatively limited public provisions for low-income families and the aged, 0 employment has persisted as the dominant source of health coverage, and employment benefits have frequently been expanded to include the full panoply of benefits usually associated with a highly developed welfare state. Although the federal government is not the direct provider of such benefits, it supports them with significant tax subsidies conditioned upon nondiscrimination rules intended, inter alia, to achieve the social welfare goal of extending coverage to all an employer's employees, not just the highly compensated.81 To illustrate, I draw on my own experience as an observer-participant in the construction and maintenance of one such highly developed employee benefit plan and in its restructuring to extend family benefits to nonmarital cohabitants. The University of California at

78 The Human Rights Campaign reports that ninety-eight state and local governments offer domestic partner health benefits. Human Rights Campaign, State and Local Governments that Offer Domestic PartnerHealth Benefits, at http://wii-hrc.org/ worknet/index.asp (last visited Feb. 22, 2001). Data on whether the benefits are extended only to same-sex partners or also to opposite-sex domestic partners are incomplete. For state and local governments for which such data are available, approximately twenty-five percent offer benefits to same-sex partners onl) the remaining seventy-five percent offer benefits to same-sex and opposite-sex domestic partners. Id State governments offering domestic partner health benefits to their employees include California, Connecticut, New York, Oregon, Vermont, and Wash- ington. Id. Major cities include Albany, Albuquerque, Atlanta, Baltimore, Chicago, Denver, Gainesville, Iowa City, Los Angeles, New Orleans, New York, Philadelphia, Phoenix, Pittsburgh, Portland, Providence, Rochester, Sacramento, San Diego, San Francisco, Santa Barbara, Seattle, and Tucson. Id. 79 See GoScnAtL, supra note 7, at 39-64. 80 These include the means-tested categorical public assistance programs that include a health care component (Medicaid), the Clinton-vra children's program for health insurance, and OASDI (Old Age, Survivors and Disability Insurance), com- monly called Social Security, with its companion Medicare program. 81 See I.R-C. § 410(b) (1994). 1284 NOTRE DAME LAW REVIEW [VOL- 76:5

Los Angeles (UCLA) employee benefits plan8 2 may rival even the Swedish welfare state at its apogee. Employees are offered a menu of health plans, starting with basic HMO plans that require no additional employee contribution. More elaborate plans are also highly subsidized and require only modest (pretax) monthly premiums. Dental and vision plans are provided at no additional cost to the employee. A richly endowed retirement plan provides, with no additional employee contribution, generous retirement and disability pensions. Low-interest loans are available for employees who wish to purchase a home. For the preschool children of employees, excellent on-site day care is available at subsidized rates. For the older children of faculty and staff, places are set aside in the University's subsidized model ele- mentary school. Exercise facilities are available at nominal charge, as is a complex of swimming pools and tennis courts intended for employees and their families. Employees and their spouses enjoy free or reduced rate tuition in academic programs. Tuition waivers for the children of employees are under consideration. There are additional self-supporting programs in which employees can benefit from group purchasing power or non-profit status, such as term life insurance, additional disability insurance, a credit union, and even a car-purchasing program. I mention only those benefits that are fully available to all employees on a non-competitive basis and exclude those that are se- lectively and competitively assigned for recruitment and retention, 8 3 such as substantial subsidies for local home purchase. At UCLA, the issue of nonmarital cohabitation first arose with health care plans and later with pension plans. The way in which the issues were discussed by the members of the local faculty welfare committee sheds light on some of the tensions and ambiguities of employee benefit plans, which are often reflected in the design of the plans themselves. Depending on the composition of the employee's family and the employee's election, UCLA health insurance is either provided for the employee, the employee and spouse, or the employee, spouse, and children. In the basic plans, each of the three options is provided without additional payment. In the plans that require some additional payment, the amount of the payment increases, but only slightly, with the number of family members covered. No

82 Some, but not all, of these benefits originate with the University of California and are system-wide. This is true of the health plan. On the other hand, other benefits, such as child care, primary education, and recreational facilities, are local. 83 These benefits run afoul of the nondiscrimination rules and thus do not receive favorable tax treatment. Most notably from the employee's perspective, they are treated like cash income and their value is thus taxed to the employee. By contrast, benefits that conform to the nondiscrimination rules are nontaxable to the employee. 2001] THE REGULARIZATION OF NONMARITAL COHABITATION 1285 distinction is made on the basis of the number of children. Ten children cost no more than one. Effectively, the level of subsidy increases sharply as the size of the family grows. The structure of employee health insurance thus seems to respond to the principle: to each employee according to his (his family's) needs. With health insurance, single persons subsidize couples and families with children, childless couples subsidize couples with children, and small families subsidize large families. These cross-subsidies are consistent ith the initial postwar decision that the employer-employee relationship, rather than the state-individual relationship, would provide the platform for the health care component of the American welfare system. In the context of health care, the differential welfare subsidies seem generally accepted, or perhaps merely unnoticed. Indeed, a subsidy for nu- merous children is common even in private, individually-purchased health insurance plans. On the other hand, there is a competing notion of equal compensation for equal service. Employee compensation includes wages and benefits.8 4 This sometimes gives rise to the claim that, because benefits form part of the compensation package, all employees should have a benefits package of equal value.85 In one sense, tie argument

84 It is frequently observed that employee benefits are part of a workers compensation package. This is true in two respects. Employment is a precondition to benefits, and net employer expenditure for benefits diminishes pro tanto the fund aa1%ilable for wage compensation. However, that granted, employee benefits resemble welfare benefits more than they do ordinary wage compensation. In the case of health coverage, for example, the value of the benefit to each employee varies according to the family circumstances of the employee and the health of the employee and his derivative family beneficiaries. Income, whether cash or in-kind, is normally subject to the personal income tax. However, the federal government exempts in-kind employee benefits from income taxation,just as it ordinarily exempts need-based welfare benefits provided by the state. (The Internal Revenue Service treats welfare payments as nontaxable income. The service takes the view that such payments are not ithin the contemplation of the I.R.C. § 61 definition of gross income. See Rev. Rul. 76-144, 1976-1 C.B. 17 (stating that disaster relief grants are not taxable income to recipients); Rev. Rul. 74-205, 1974-1 G.B. 20 (stating that the IRS has consistently held that payments made under legislatively provided social benefit programs for promotion of the general welfare are not includible in a recipient's gross income); Rev. Rul. 63-136, 1963-2 C.B. 12 (stating that cash benefits paid to persons undergoing on-the-job train- ing in the Federal Manpower program are not taxable income to the recipients); Rev. Rul. 57-102, 1957-1 C.B. 26 (stating that state welfare benefits for the blind are not includible in the gross income of the recipients for federal income tax purposes).) Given the lack of universal health coverage in the United States, the direction of this tax subsidy is troubling. Effectively, those who do not have employer-provided group health insurance subsidize those who do. 85 See, e.g.,James E. Donovan, An EthicalArgumentto Restri Domestic Partnershipsto Same-Sex Coupes, 8 LAw & SFxuAur 649, 666 n.63 (1998) (reporting an e-mail com- 1286 NOTRE DAME LAW REVIEW (VOL, 76:5r seems questionable, for its proponents would not ordinarily assert that all employees should receive equal wages, without regard to the value of their services. Thus, the very premise of the claim, that all benefit packages should have equal value regardless of the relative value of each employee's services, suggests that benefits are different in character than wages, that is, that they are intended to serve welfare goals. This view is reinforced by the generous federal tax subsidies for employee benefit plans and the nondiscrimination requirements for those subsidies, which require that highly compensated employees not be preferred in an employer's benefit scheme.8 6 Moreover, the federal government exempts qualifying in-kind employee benefits from personal income taxation, 7 just as it ordinarily exempts need-based welfare benefits provided by the state. 8 Nevertheless, the claim to a benefits package of equal value has had some valence. It is expressed in the notion of a "menu" of benefits from which an employee can choose according to his needs and preferences, with the implicit suggestion, if not the assurance, that any particular set of choices is equal in value to any other set of choices. In health insurance, the claim to equality of benefits may implicitly be expressed in those employer-provided health plans that insure the worker alone and require the worker to contribute a full unsubsidized premium for any other family members he wishes to cover. Although there are other explanations for such a plan, including employer cost- munication in which Suzanne Goldberg, staff attorney for Lamda Legal Defense and Education Fund (LLDEF), explained that LLDEF supports inclusion of opposite-sex, as well as same-sex, couples in domestic partnership plans on the ground that these are largely -economic issues of equal compensation). Alternatively, employees might claim that workers with equal wages should have equal benefits; in other words, a worker's benefit package should correspond in value to his wages. In other words, higher-wage employees should have better benefit packages. Federal anti-discrimination rules stifle this claim and, in doing so, shed light on the role that benefits play in the national welfare scheme. 86 See I.RC. § 410(b) (1994). 87 However, the exemption from income taxation does not extend equally to benefits provided for a domestic partner, as opposed to a spouse, of an employee. Under federal income tax law, married employees may exclude spousal coverage from income. Unmarried employees may only exclude the value of partner coverage if the partner is a dependent of the employee. I.R.C. §§ 105(B), 106, 152 (1994 & Supp. V 1999); see also Lindsay Brooke King, Enforcing ConventionalMorality Through Taxation? Determining the Excludability of Employer-Provided Domestic PartnerBenefits Under Sections 105(B) and 106 of the Internal Revenue Code, 53 WAsm. & LEE L. REV. 301, 303 (1996); William V. Vetter, Restrictions on Equal Treatment of Unmarried Domestic Partners,5 BU. PUB. INT. L.J. 1, 5-12 (1995). 88 See sources cited supra note 84. 2001] THE REGULARIZATION OF NONMIARITAL COHABITATION 1287

cutting, they nevertheless express the principle that the employer is subsidizing each employee's health insurance in equal measure. In UCLA Faculty Welfare Committee deliberations, the "equality of value" argument has recurrently been made, for example, ith respect to proposals to expand the provision of subsidized University day care and to introduce University of California tuition waivers for the children of employees. "Why," childless faculty ask, "should I be required to subsidize the care and education of other employees' children?" 89 The answer, perhaps unsatisfactory to those who make the equality-of-value argument, is that benefits are essentially welfare allo- cations distributed on the basis of need. Under the welfare allocation view, all employees, those who use a benefit such as day care and those who do not, are ultimately equally situated because the non-users are in any event wealthier in that they do not have to devote any portion of their income to the purchase of (partially subsidized) child care. Similarly, those whose poor health requires that they make extensive use of their health coverage derive disproportionate value from that coverage, but are ultimately no better off in terms of disposable wealth than their healthier co-workers. Thus, the comprehensiveness of the UCLA welfare benefits system tends to insure that all employees have a substantially equal opportunity to enjoy the fruits of their Awage income, and the salary scale for that wage income is, contradistinctively, set to compensate each employee for the value of services rendered to the University. That employee benefits may simultaneously be interpreted by employees as in-kind equivalent for wage compensation and as employment-based social welfare has significance for the extension of benefits to nonmarital cohabitants and for any analysis of the reci- procity of rights and responsibilities. If benefits are predominantly characterized in terms of their social welfare function, then the claim of employees and their nonmarital partners to benefits is properly based on the assertion that nonmarital partners are equally situated with marital partners. That assertion was initially made with respect to family health benefits by gay and lesbian employees, who complained that they maintained households indistinguishable from marital households, they would marry their partner if state law allowed them to do so, and therefore their employer should treat them as though they were married. Their claim to family health benefits was compel-

89 And then: "Why have the benefit at all? Why not let the market provide?" (Of course, the follow-up objections would not be made about health care, from which all stand to benefit, so I assume that the gravamen of these follow-up arguments is simply: "I don't stand to benefit at all.") 1288 NOTRE DAME LAW REVIEW [VOL. 76:5 ling in view of the state law restriction of marriage to opposite-sex couples. In the course of discussion, the UCLA Faculty Welfare Com- mittee debated the inclusion of nonmarital opposite-sex couples. The Committee concluded that opposite-sex couples should not be in- cluded, because they could marry if they wished and thereby undertake marital obligations to each other. Their failure to do so undermined their claim to equality of treatment. In other words, they should not be allowed to claim the benefits of marriage without as- suming the responsibilities of marriage. 9° The Regents of the Univer- sity of California apparently agreed on this point 9 ' when, in 1997, they 9 2 extended family health benefits to same-sex domestic partners only. For this purpose, same-sex domestic partners are required to join to-

90 Most notably in California, both spouses have "present, existing, and equal interests" in the marital earnings of either spouse. CAL. FAm. CODE § 751 (West 1994). Subsequently, the system-wide University of California Faculty Welfare Commit- tee tried to piggy-back on the grant of health benefits to same-sex partners to obtain further extension to opposite-sex partners. See UNIv. COMM. ON FACULTY WELFARE, ACADEMIC SENATE, UNIV. OF CAL., ENSURING FULL EQUALITY IN BENEFITS FOR UC EMI. PLOYEES WITH DoMEsnc PARTNERS 3 (1999) [hereinafter ENSURING FULL EQUALr1Y]. Failure to include the latter, it implied, is a form of prohibited discrimination on the basis of sexual orientation and marital status. In other words, extension of health benefits only to same-sex partners impermissibly discriminates against opposite-sex partners on the basis of sexual orientation. And, allowing benefits for married persons but not opposite-sex cohabitants discriminates against the latter on the basis of marital status. Essentially, the argument was that the declaration required of same-sex couples should also suffice for opposite-sex couples, and it is immaterial (i) that the former cannot marry and the latter can and (ii) that the declaration required of same-sex domestic partners encompasses rights and responsibilities far more limited than those imposed by the law of marriage. Id. 91 1 leave aside the instrumental issues of recruitment and retention. Injustifying their extension of health coverage to the same-sex partners of University of California employees to Republican Governor Pete Wilson, who strongly opposed the move and had appointed many of them, the Regents buttressed their equality rationale with the need to compete with other leading universities that were already offering benefits to same-sex partners. Recruitment and retention surely were issues for a state university system that was somewhat late to join the trend. Moreover, the tight labor market of the prosperous 1990s generally gave employers strong incentive to enhance (generally noninflationary) employee benefits as a means of attracting new employees. The tight labor market may explain the rapid spread of same-sex partner benefits, but it does not explain their genesis or their particular rationales. 92 The Academic Council, an advisory body to the Regents, took the opposite position, recommending that the University of California provide health benefits to opposite-sex partners as well. See ENSURING FULL EQUAITy, supra note 90. The Regents extended health benefits to same-sex domestic partners on the same terms that they offered to the families of married employees, that is to partners and the dependent children of either partner. 2001] THE REGULARIZATION OF NONMARITAL COHABITATION 1289 gether to execute the following affidavit (and may be required to produce evidence of its truth 93): 1. We are each other's sole domestic partner and intend to remain so indefinitely. We are in a relationship of mutual support, car- ing, and commitment. We are financially interdependent. 2. We are of the same sex, neither one of us is legally married, and we are not related by blood to a degree of closeness which would prohibit legal marriage in the State of California. 3. We are at least 18 years of age and have the capacity to enter a contract. 4. We have resided together for at least six months and intend to reside together indefinitely. 5. It has been at least six months since the termination of a previ- 94 " ous same-sex domestic partnership. At the same time, the Regents also provided that an employee who enrolls neither a spouse nor a same-sex domestic partner may enroll an adult dependent relative who (i) resides with the employee, (ii) is incapable under California law of marriage with the employee because of the consanguineous relationship, and (iii) is claimed by the employee as a tax dependent. 95 This provision may also be understood as an equality-based welfare provision in that an employee who resides with and assumes responsibility for an adult relative is similarly situated, for health benefit purposes, to an employee who resides ith a spouse or same-sex domestic partner. Although die employee is not required to undertake any further obligation to the dependent relative, the moral obligation to care for family members, already made manifest by the employee having taken de facto responsibility for the dependent relative, serves as an effective substitute for marriage, in the case of spouses, or the declaration and proof of domestic partnership, in the case of same-sex nonmarital cohabitants.

93 An employee claiming same-sex partner health benefits must submit upon re- quest proof of at least three of the following items: joint mortgage or joint tenancy on a residential lease; joint bank account; joint liabilities (e.g., credit cards or car loans);joint owmership of significant property (e.g., cars); durable property or health care powers of attorney; wills, life insurance policies or retirement annuities naming each other as primary beneficiary; written agreements or contracts showing mutual support obligations or joint ownership of assets acquired during the relationship. UNIV. OF CAL., AFFADAVrr OF SAiME-SFx DoNmisrc PARTNERSHip OR ADULT DE.PEXtDL\-r RELATIVE RELATIONsHIP 1-2 (2000), available at http://-ww.ucop.edtt/bencom/ forms/forms.htiml (last modified July 3, 2001). 94 IM 95 Id- 1290 NOTRE DAME LAW REVIEW [VOL- 76:5

On the other hand, the eligibility structure also suggests some concern about providing an "equal benefits package." From a welfare perspective one might, for example, allow an employee to claim both a spouse and another adult dependent relative residing with the employee. The University of California plan allows the possibility that some married employees, whose spouses are otherwise covered and thus do not require the employee's health insurance, might wish instead to insure an adult dependent relative. An employee may do so, but only if he dis-enrolls his spouse. In other words, an employee is allowed to enroll only one adult beneficiary in the health plan, which does suggest concern about providing an equal benefits package to each employee. The rights and responsibilities issue was presented much more sharply in the claim (as yet unresolved) that same-sex domestic partners should enjoy the UCLA subsidy for joint-and-survivor pension benefits currently available only to married employees and their spouses. 96 An employee may name any contingent annuitant, but ordinarily only at the cost of full actuarial reduction of the employee's lifetime benefit.97 However, spousal contingent annuitants receive a substantial subsidy that, depending on the plan, completely or partially eliminates actuarial reduction of the married employee's life- 98 time benefit. For family law purposes, health coverage and pensions are mark- edly different in character. When a married employee divorces, his former spouse is no longer eligible for employer-subsidized health coverage. 99 Although a community, or marital, asset during marriage, health insurance has no distributable value at divorce. By the terms of the policy, the former spouse is excluded from coverage, and the employee continues to fund his now separate health insurance with his post-divorce separate wages. Correspondingly, a same-sex domestic partner's entitlement to health insurance ends at the termination of the relationship with an employee. Thus, the extension of health in-

96 UNIV. OF CAL., RETIREMENT PLAN SUMMARY PLAN DEscaunoN 14-20 (2000), available at http://www.ucop.edu/bencom/rs/ucrp.html (last modified July 27, 2001). 97 Id. at 17-18. 98 Id. 99 Federal law provides short-term relief for persons who have lost eligibility for employer-provided group health insurance on account of certain qualifying events, including divorce. The Consolidated Omnibus Budget Amendment Act of 1985 (CO- BRA) requires that the sponsors of ERISA-regulated group health insurance plans offer an employee's divorced spouse thirty-six months of individual health insurance continuation. 29 U.S.C. §§ 1161-1168 (1994). The sponsor may charge the divorced spouse the full premium. Id. 2001] THE REGULARIZATION OF NONMARITAL COHABITATION 12 91

surance coverage does not raise marital or partnership termination issues. In contrast to health insurance, which provides current coverage, pension contributions provide only deferred benefits. For this reason, pensions are frequently described as "deferred compensation," and marital property law classifies them as marital or separate property according to when they were earned, not according to when benefits are received. 100 Thus, in every American jurisdiction, divorcing spouses have marital property claims to pension rights earned by either spouse during marriage. (In California, spouses are entitled to an equal division of the community property. 01) Therefore, married University of California employees not only have greater earned pension rights, but also have an obligation to share pension rights earned during marriage with a spouse in the event of divorce, a significant obligation in a state in which a marriage is equally likely to end in divorce as in death. (And in the case of nonmarital cohabitation, the odds of pre-death dissolution are likely to be even greater.) This raises the question whether employees can plausibly demand for themselves and their same-sex domestic partners the subsidized joint- and-survivor pensions enjoyed by married employees without also agreeing to shoulder the obligation to share pension rights earned during their relationship with a partner should the domestic partnership end before death. The "equal value of employee benefits" argument might be understood to support a "rights, but no responsibilities" claim by the employee who has a same-sex partner. In other words, "I am entitled to a joint-and-survivor pension benefit of equal dollar value to my married co-worker even though I have no post-termination responsibility to share pension rights earned during our relationship with my domestic partner." The welfare analysis would suggest, however, that the same- sex claimant is entitled only if he is equally situated with respect to pension benefits. He may arguably be during the continuance of the

100 For a national survey, see Grace Ganz Blumberg, MaritalPropery Treatment of Pensions, Disability Pay, Workes' Compensation, and Other Wage Substitutes: An Insurance, or Replacement, Analysis, 33 UCLA L. REv. 1250, 1256-66, 1290-98 (1986). When the value of the pension is substantially related to the number of years of service, a "time rule" is usually applied to determine the fractional marital interest in the pension. The numerator is the number of years of service during the marriage, and the denominator is the total number of years of service that contribute to the value of the pension. See e.g., In re Marriage of Gowan, 62 Cal. Rptr. 2d 453, 456-59 (Ct. App. 1997); In reMarriage of Poppe, 158 Cal. Rptr. 500, 503-04 (Ct. App. 1979). 101 See CA. FA _. CODE § 2550 (West 1994) (providing that "the court shall... divide the community estate of the parties equally"). 1292 NOTRE DAME LAW REVIEW [VOL- 76:5 domestic partnership, when, in the course of ordinary domestic life, he shares his income with his partner, but he is not at the inter vivos dissolution of the partnership. The University of California is not generally in the business of determining how property interests should be distributed at the dissolution of a same-sex domestic partnership. (On the other hand, it would not be a great extension of the same-sex domestic partnership affidavit to require, as a prerequisite to the pension subsidy enjoyed by married persons, that domestic partners agree that the non-employee partner may claim a quasi-marital domestic partnership division of the pension at the termination of their relationship.) To the extent that the University declines to provide a scheme for pension distribution at the termination of domestic partnerships and also understands the marriage subsidy as a welfare subsidy reflecting the rights and obligations of marriage, the absence of any obligation to share at termination may undermine the claim of domestic partners to the benefits they seek during their ongoing relationship.

III. THE INTER SE OBLIGATIONS (BURDENS) OF COHABITATION

The preceding Parts of this Essay involve claims that a couple may make against the state or one party's employer. In contrast, this Part treats claims that one party may make against the other at the termination of their relationship. This is the direct province of Marvin. Under what circumstances should such claims be recognized and, implicitly, under what circumstances should they be rejected? Marvin adopted the rubric of contract: inter se claims should be recognized only to the extent both parties agreed they would be. Marvin has been much criticized as unworkable, inapt, artificial, and inadequately responsive to a range of worthy claims.'02 It is not my purpose here to review that criticism, but rather to discuss an alternative approach and to situate that approach in a body of law and practice that not only recognizes marriagedike obligations, but also, as this Essay has shown, increasingly recognizes marriage-like rights in employment benefits and in civil status.

102 See ALI PRINciPLEs (Tentative Draft 2000), supra note 24, § 6.03, reporter's n, cmt. c; Blumberg, supra note 2, at 1160-70; sources cited supra note 4. Application of the rubric of contract to rfonmarital cohabitation has generated considerable dissatisfaction, for it tends to produce two problems. Either courts reach harsh and undesirable results by applying contract law strictly. Or, in an effort to avoid harsh results, courts play havoc with contract law, distending it'beyond recognition. Moreover, the contractual rubric tends to be very difficult and time-consum- ing to administer. 2001] THE REGULARIZATION OF NONMARITAL COHABITATION 1293

A. Status Developments in American State Law

Since Marvin was decided, a few state courts have taken the path that the California Supreme Court declined to take in Marvin when it disapproved a brief line of decisions of the California Court of Ap- peal, including In re Marriageof Cay,103 which applied the California Family Law Act104 to nonmarital partners who lived in stable nonmarital cohabitation. Caty effectively asked: "Has this nonmarital family behaved like a marital family? If so, why not apply our already well-developed family law?" With respect to property rights, Cary has been followed by Washington, which applies its state community prop- 0° erty law to long-term stable nonmarital domestic relationships.1 Mis- sissippi has occasionally done likewise,10 6 and several Oregon

103 109 Cal. Rptr. 862, 866-67 (Ct. App. 1973). To the same effect, see Estate of Atherley, 119 Cal. Rptr. 41, 46-48 (Ct. App. 1975). But see Beckman v. Mayhew, 122 Cal. Rptr. 604, 607-08 (Ct. App. 1975) (rejecting the holding and anal)sis of Caryand Atherley). 104 CAL. CiV. CODE §§ 4000-5137 (West 1994) (repealed 1992). 105 See Connell v. Francisco, 898 P.2d 831, 835-37 (Wash. 1995) (holding that at the dissolution of "meretricious" relationships, the court should rely on the community property laws to divide between the partners all property that would have been community property had they been married). Conner provides guidance in identify- ing "meretricious relationships" for the purpose of this rule. Calling the phrase "a term of art," it described a "meretricious relationship" as a stable, marital-like relationship where both parties cohabit with knowledge that a lawful marriage between them does not exist. ... Relevant factors establishing a meretricious relationship include, but are not limited to: continuous cohabitation, duration of the relationship, purpose of the relationship, pooling of resources and services for joint projects, and the intent of the parties.... [A] relationship need not be "long term" to be characterized as a meretricious relationship.... While a "long term" relationship is not a threshold requirement, duration is a significant factor. A "short term" relationship may be characterized as meretricious, but a number of significant and substantial factors must be present. Id.at 834 (citations omitted); see also Marriage of Lindsey, 678 P.2d 328, 331 (Wash. 1984) ("[C]ourts must 'examine the [meretricious] relationship and the property ac- cumulations and make ajust and equitable disposition of the property." (alteration in original) (quoting Latham v. Hennessey, 554 P.2d 1057, 1059 (Wash. 1976))); Foster v. Thilges, 812 P.2d 523, 526 (Wash. Ct. App. 1991) (holding that the court need not resolve the parties' conflicting claims regarding their intentions in acquiring property, because "[w]here the relationship was long-term, stable, pseudomarital and the undertakings were joint projects as in the instant case ....the couple's property is to be divided justly and equitably, applying community property principles"). 106 See eg., Pickens v. Pickens, 490 So. 2d 872, 875-76 (Miss. 1986) (holding that a homemaker has an equitable claim to property accumulated during a long-term, cohabiting relationship, without regard to a contract inquiry); se also Evans v. Wall, 542 So. 2d 1055, 1056-57 (Fla. Dist. Ct. App. 1989) (imposing a constructive trust or, 1294 NOTRE DAME LAW REVIEW (V€OL- 76:5 intermediate appellate decisions have taken a somewhat more equivo- cal status approach to long-term cohabitation. 10 7 Courts in other states nominally follow the contract rubric of Marvin, but nevertheless make awards that are difficult to justify under the law of contract.1 0 To the extent that the law of contract does not reach such results, these courts also effectively apply a status construct, that is, they reach a result that seems equitable in light of the nature of the parties' family relationship. Finally, courts that have been willing to infer an agreement from the parties' domestic behavior'0 9 straddle the bound- ary between contract and status. Conduct that provides the basis for inferring an agreement is generally also susceptible to a status treatment. For example, the parties' assumption of complementary and cooperative family roles can support a finding of implied agreement to share property acquired during the relationship, or, more directly, support the imposition of a status-based obligation to share property.110 alternatively, an equitable lien to allow a woman to recover for contributions made to improve her former partner's land); Sullivan v. Rooney, 533 N.E.2d 1372, 1374 (Mass, 1989) (relying upon constructive trust doctrine to award a woman a one-half interest in the home in which the partners lived during the cohabitation, but which was titled solely in the man's name). 107 See Wilbur v. DeLapp, 850 P.2d 1151, 1153 (Or. Ct. App. 1993) ("[We) may distribute property owned by the parties in a non-marital domestic relationship. The primary consideration in distributing such property is the intent of the parties. How- ever, ... in distributing the property of a domestic relationship, we are not precluded from exercising our equitable powers to reach a fair result based on the circumstances of each case." (citations omitted)); Shuraleff v. Donnelly, 817 P.2d 764, 768-69 (Or. Ct. App. 1991) (involving a fourteen-year cohabitation). 108 See, e.g., Kozlowski v. Kozlowski, 403 A.2d 902, 906 (NJ. 1979). But see Watts v. Watts, 448 N.W.2d 292, 296-97 (Wis. Ct. App. 1989); Ira Mark Ellman, The Theoy of Alimony, 77 CAL. L. REv. 1, 21-23 (1989). 109 Marvin contemplates agreements implied by the parties' domestic behavior, and some California courts have been willing to infer agreements from the parties' domestic behavior. See, e.g., Byrne v. Laura, 60 Cal. Rptr. 3d 908, 914 (Ct. App. 1997); see also Glasgo v. Glasgo, 410 N.E.2d 1325, 1331 (Ind. Ct. App. 1980) ("Recovery for parties seeking relief would be based only upon legally viable contractual and/or equitable grounds which the parties could establish according to their own particular circumstances."), quoted with approval in Bright v. Kuehl, 650 N.E.2d 311, 315 (Ind. Ct, App. 1995); Goode v. Goode, 396 S.E.2d 430, 435-38 (W. Va. 1990) (allowing recovery based on express or implied contract); Watts v. Watts, 405 N.W.2d 303, 311-12 (Wis. 1987). 110 Consider, for example, how the facts of In re Marriage of Cary, 109 Cal. Rptr. 862, 863 (Ct. App. 1973), would have been treated by the Marvin court under its expansive contract rubric. The complementary behavior of the parties in having and raising four children provided ample basis for an implied "share and share alike" contract. M 2001) THE REGULARIZATION OF NONMARITAL COHABITATION 1293

B. The American Law Institute's Treatment of the Inter Se Obligations of Nonmarital Cohabitants

Chapter 6 of the ALI Principles of the Law of Family Dissolution"1 treats the financial claims that one nonmarital partner may have against the other at the termination of their relationship.' 12 Chapter 6 takes the view that the equitable considerations shaping the financial rules that the Principlesapply at the end of a marriage are equally pertinent at the end of a stable nonmarital cohabitation of substantial duration, whether same-sex or opposite-sex.113 Thus, at the termination of nonmarital cohabitation Chapter 6 applies, with minor varia- tion, the same rules that control property distribution and continuing obligations, if any, to share income after a marriage has ended. Chap- ter 6 was presented to the ALI membership for final approval in May 2000.114 After vigorous debate, it was approved by a substantial major- ity of the members. The ALI treatment begins with the premise that, if both parties so desire, they may always join together and explicitly contract for the terms of their relationship.' 1 5 Thus, the task of Chapter 6 is to formulate an appropriate set of default rules, in other words, the rules that should govern when the parties have not otherwise provided. Chapter 6 could have required that a nonmarital claimant affirmatively prove an agreement specifying the claims available at permanent separation. Or, it could have provided equitable rules that would apply to qualifying domestic relationships unless one of the parties proves an inconsistent agreement. The ALI chose the latter approach for many reasons, which are extensively discussed in the commentary of Chap-

111 ALI Pm capis (Tentative Draft 2000), supra note 24, § 6.01. The ensuing discussion of the ALI Pincipesis adapted from prefatory materials prepared by the au- thor for Chapter 6 of the Principles. Those prefatory materials will be published in Chapter 1 of the ALI Principles. 112 Chapter 6 is limited to the following question: what are the economic rights and responsibilities of the parties to eadi othwr at the termination of their nonmanital cohabitation? The Prindplesdeal onlywith the obligations of the parties to each other and to their children at family dissolution. They do not create any rights against the government or third parties. Thus, Chapter 6 should not be understood to reive ie doctrine of common-law marriage. 113 Chapter 4 regulates property distribution, and Chapter 5 recasts spousal support as "compensatory payments." See PRnciPLrs OF THE LxW OF FAMILY DssoLUtON: ANALYsis AND RECOMNMENDATIONS chs. 4, 5 (Proposed Final Draft 1997) [hereinafter ALI Pnmci.s (Proposed Final Draft 1997)]. 114 ALI PRnwcipL.Vs (Tentative Draft 2000), supra note 24, cl. 6. 115 Chapter 7 of the Prindplesallowsthem to do so, within the formal and substantive limitations set out by that chapter for contracts between marital and nonmariul partners. 1296 NOTRE DAME LAW REVIEW [VOL- 76:5t ter 6.116 Those reasons roughly fall into two categories. The first re- lates to the intentions of the parties. The second is normative. The ALI concluded that the parties' failure to marry should not be understood to signify that the parties agreed that they would have no economic obligations to each other. 1 7 First, as the incidence of cohabitation has dramatically increased and cohabitation has become socially acceptable at all levels of society, it has become increasingly implausible to attribute special significance to the parties' failure to marry. Jan Trost, the Swedish sociologist, first observed this in Swe- den and Denmark, which have long had the highest cohabitation rates in the world. Trost concluded that the more frequent cohabitation becomes, the less cohabitation signifies. 118 When cohabitation becomes well established as a social institution, people do not choose to cohabit rather than marry. They simply cohabit. 1 9 Second, domestic partners may fail to marry for many reasons. Among others, some may-have been unhappy in prior marriages and therefore wish to avoid the form of marriage, even as they enjoy its substance with a domestic partner. Some begin a casual relationship that develops into a durable union, by which time a formal marriage ceremony may seem awkward or even unnecessary in view of widespread, albeit generally erroneous, popular belief that time alone transforms cohabitation into common-law marriage.' 20 Failure to marry may reflect group mores. Some ethnic and social groups have a substantially lower incidence of marriage and a substantially higher incidence of

116 See ALI PRINcIPLEs (Tentative Draft 2000), supra note 24, §§ 6.02 cmts., 6.03 cmts. 117 In the debate that preceded the ALI membership's adoption of Chapter 6, the crux of the opposition position seemed to be that the failure of cohabitants to marry should be interpreted to signify an agreement of the parties that they would have no obligations to each other should their relationship end. Thus, the adoption of Chap- ter 6 would frustrate the intentions of both parties to a cohabiting relationship. The ALI position on the intentions of the parties with respect to the possible demise of their relationship is that nonmarital couples who do not make explicit contracts about the subject are most likely to either have no intentions at all or have no common intentions. It is of course possible that one of the parties may avoid marriage with the hope of avoiding obligation to the other, but the ALI concluded that the law has no interest in vindicating the intention of one party only. Under the ALl Principles, that party must secure the consent of the other to avoid any duties that would otherwise flow from the relationship. 118 JAN TRosT, UNMARRIED CoHABrrATION 185-87 (1979). 119 For more recent discussion of this point, see Kathleen Kiernan, The Rise of Cohabitation and Childbearing Outside Marriagein Western Europe, 15 INT'L J.L. POL' & F.fi. 1, 3 (2001), and sources cited therein. 120 In the United States, this is expressed in the popular myth that, after seven years of cohabitation, cohabitants are common-law spouses. 2001] THE REGULARIZATION OF NONMARITAL COHABITATION 1297 informal domestic relationships than others. Failure to marry may also reflect strong social or economic inequality between the partners, which allows the stronger partner to resist the weaker partner's preference for marriage. Finally, there are domestic partners who are not allowed to marry each other under state law because they are of the same sex, although they are otherwise eligible to marry and would marry one another if the law allowed them to do so. In all of these cases, the absence of formal marriage may have little or no bearing on the character of the parties' domestic relationship and on the equitable considerations that underlie claims between laxful spouses at the dissolution of a marriage. Normatively, the ALl takes the position that family law should be concerned about relationships that are indistinguishable from marriage, except for the legal formality of marriage. The more frequent such relationships become, the more the law should be concerned. Chapter 6 assumes as its foundation Chapters 4 and 5 of the Prindples, which define and rationalize the economic claims that one spouse has upon another at the termination of a marriage. The equitable concerns that are expressed there apply equally to marriage-like cohabitation. Chapters 4 and 5 do not conceive marriage as a contract whose terms require the equitable remedies contained in those chapters. They instead see those remedies and obligations as arising over time 2 1 from the parties' conduct in sharing their lives. The ALl sought to formulate rules that would distinguish relationships that are marriage-like from those that are not. It also sought rules that would be readily administrable, but would allow for individualized treatment, in other words, rules that would easily dispose of run-of-the-mill cases but that would allow persons to show that they should not be covered by the rules. Finally, it sought rules that would give due weight to the economic interests of children born in nonmarital relationships. In general, the ALI defines domestic partners as "two persons of the same or opposite sex, not married to one another, who for a significant period of time share a primary residence and a life together as a couple." 22 Routine cases do not, however, require direct proof that the relationship of the parties fully satisfies this definition. In- stead, the operative provisions of the Principlesset out an absolute rule

121 A similar approach has been entertained by the Reporter of the Uniform Pro- bate Code. See Lawrence W. Waggoner, Marital Property Rights in Traisition,59 Mo. L REv. 21, 36-41 (1994) (proposing an intestacy statute for unmarried couples who lived in a "marriage-like relationship" prior to the death of one of them). 122 ALI PanwcipLs (Tentative Draft 2000), supra note 24, § 6.03(1). 1298 NOTRE DAME LAW REVIEW [VOL. 76:5

and a presumption, which assure that few cases will involve any bur- densome factual inquiry. When parties have lived together in a common household for a specified uniform period of time with a child of both of them, they are domestic partners. 123 The ALl Principlesdo not specify any particular period of time, but leave that decision to the enacting state. 124 (Other English-speaking jurisdictions often set a two- or three-year minimum 2 5 period of cohabitation. ) When parties are not the co-parents of a child, but have shared a common household for a specified uniform period of time, they are presumed to be domestic partners. 26 The specified period should be longer when they do not have a child. For example, if two years were used when the parties do have a child together, three years would be appropriate when they do not. The presumption requires that the parties share a common household, which is generally defined as one that is occupied exclusively by the parties and their relatives.' 27 The purpose of this requirement is to make the presumption unavailable to persons who merely share a group residence, such as college stu- dents. 28 The presumption may be rebutted by a showing that the parties did not in fact share life together as a couple, as that term is defined in the guidelines for determining whether parties shared life together as a couple. 2 9 If neither the absolute rule nor the rebutta- ble presumption applies, the claimant bears the burden of showing that (i) for a significant period of time (ii) the parties shared a primary residence and (iii) a life together as a couple within the meaning of the guidelines. 30 Such cases should be rare. Under the contractual rubric of Marvin, every cohabitation case requires extensive fact finding. By contrast, the ALl formulation requires such fact finding only when a party seeks to rebut the presumption that the parties are domestic partners on the ground that they did not in fact share life together as a couple, or a claimant who cannot bring himself within the absolute rule or the presumption nevertheless seeks to assert a claim.' 3 1 Such efforts are likely to be rare because most would be either unsustainable or of little economic

123 Id. § 6.03(3)-(4). 124 Id § 6.03(6). 125 See infra text accompanying notes 137-56. 126 ALI PRINCIPLES (Tentative Draft 2000), supra note 24, § 6.03(3)-(4). 127 See id. § 6.03(4). 128 See id. § 6.03 illus. (i). 129 See id. § 6.03(7). 130 Id. § 6.03(6). 131 See id. § 6.03. 2001] THE REGULARIZATION OF NONMARITAL COHABITATION 1299 value. For those infrequent cases, the ALl sets out criteria intended to guide the court in determining whether the parties shared life together as a couple for a significant period of time. That list includes, inter alia, the character of the parties' social and economic relationship, the extent to which the relationship wrought changes in the life of either or both parties, the parties' community reputation as a couple, and the extent to which the parties acknowledged commitment and responsibilities to each other, including the parties' enroll- ment as domestic partners in a benefit scheme or employee benefit 2 plan.' 3 Once the parties have been found to be domestic partners, they are generally subject to the property and compensatory payment (spousal support)' 3 3 rules of the ALI Principles,unless they have made an inconsistent agreement setting different terms. 34 With respect to compensatory payments, only relatively long-term cohabitation can give rise to compensatory claims, for they have a minimum vesting period.' 35 Domestic partnership status alone is insufficient to generate significant compensatory payment claims.' 3 6 Similarly, the amount of marital property is usually proportionate to the duration of the marriage. Thus, the duration of cohabitation is likely to be the main determinant of property claims. The most troubling aspect of the Marvin contractual rubric has been its incapacity to respond to long-term cohabitation claims. 137 By contrast, those are the claims most likely to be reached effectively by the ALI rubric.

C. The Use of Status Constructs by Our Cultural Cousins

In developing Chapter 6, the ALI reporters did not have to search for models in exotic or remote places. In addition to the states of Washington' 38 and Oregon,139 the reporters received guidance

132 See id § 6.03 (7) (a)-(m). 133 Chapter 5 of the ALI Principles rationalizes and reconceptualizes spousal support as "compensatory payments." ALI PRirNcL.cS (Proposed Final Draft 1997), supra note 113, § 5.01 cmt. a. 134 ALI PINcIPLEs (Tentative Draft 2000), supranote 24, § 6.01(2). 135 ALI PmNciPLs (Proposed Final Draft 1997), supra note 113, § 5.05 cmt. c. 136 Id. 137 See ALI PRuNCIPLEs (Tentative Draft 2000), supra note 24, § 6.03, Reporter's Note cmt. c. 138 See supra note 105. 139 See supra note 107. 1300 NOTRE DAME LAW REVIEW [VOL- 7625 from our closest English-speaking cultural cousins, particularly Ca- 140 nada, Australia, and New Zealand. Ontario (Canada) includes cohabitants as well as lawfully married persons in its statutory definition of "spouse" for purposes of inter se support obligations. 141 A "spouse" includes "either of a man and woman who are not married to each other and have cohabited, (a) con- tinuously for a period of not less than three years, or (b) in a relationship of some permanence, if they are the natural or adoptive 142 parents of a child." In M. v. H., 14 3 a same-sex cohabitant challenged the Ontario statutory definition ("either of a man and woman who are not married to each other and have cohabited") as violative of the equal protection guarantee of the Canadian Charter.144 Reasoning that financial de- pendency and interdependency, to which the statute is designed to respond, are no less likely in same-sex cohabitation than in opposite- sex cohabitation, the Supreme Court of Canada concluded that the exclusion of same-sex cohabitants is not rationally connected to the legislative objective of assuring adequate economic support for cohabitants at the termination of their relationships. 45 Effectively combining Ontario statutory and case law, the British Columbia (Canada) support statute includes both opposite-sex and same-sex partners within its definition of "spouse," providing that, for purposes of spousal support, a "spouse" includes a person who "lived with another person in a marriage-like relationship for a period of at

140 Indeed, shortly before Chapter 6 of the Principles was presented to the ALI membership in May 2000, the government of New Zealand introduced legislation closely resembling Chapter 6. See Vernon Small, Married and Gay Equal in New Law, N.Z. HERALD, Apr. 3, 2000, at 1. For current Canadian, Australian, and New Zealand legislation, see text immediately infra. Early Canadian and Australian legislation is discussed by Carol Bruch, Nonmarital Cohabitationin the Common Law Countries:A Study inJudicial-LegislativeInter- action, 29 AM. J. Comw. L. 217, 234-35 (1981). For a survey of Latin-American, Euro- pean, and Israeli treatment of stable, nonmarital cohabitation in the early 1980s, see Blumberg, supra note 2, at 1170-76. The first issue of 15 INrr'LJ.L. POL'Y & F.Am. (2001) is devoted entirely to articles on the demography and legislative regulation of cohabitation in Europe. 141 Ontario Family Law Reform Act of 1986, R.S.O., ch. F-3, §§ 29-30 (1990) (Can.). 142 id. An earlier version of the Act required, under subsection (a), continuous cohabitation for not less than five years. Ontario Family Law Reform Act of 1979, eh. 2, § 14(b), 1978 S.O. 5 (Can.). 143 [1999] D.L.R. 577. 144 Id. at 578. The Canadian Charter is equivalent to the United States Constitution. 145 I& at 581. 2001] THE REGULARIZATION OF NONMARITAL COHABITATION 1301 least 2 years... and, for the purposes of this Act, the marriage-like relationship may be between persons of the same gender." 40 Legislation of New South Wales (Australia) defines a de facto relationship as a "relationship [same-sex or opposite-sex] between two adult persons... who live together as a couple, and... who are not married to one another or related by family."147 The legislation includes a list of criteria to determine whether a relationship constitutes a de facto relationship. 48 De facto relationships give rise to support and property rights at the termination of the relationship by inter vivos separation, as well as by death.149 The law of New South Wales also creates quasi-marital rights against third parties and the state. 50 In April 2000, the government of New Zealand rejected a New Zealand Law Commission proposal that it create a special "domestic partnership" status for the one out of seven New Zealand couples, same-sex and opposite-sex, who live in nonmarital cohabitation.' 5 1 In- stead, the government proposed legislation that would treat long-term (more than three years) stable nonmarital cohabitation equally with marriage. 152 The proposed legislation would bring such couples, same-sex and opposite-sex, under the New Zealand Marital Property Act, which establishes the parties' rights at death and inter vivos dissolution.153 At dissolution, the Act requires an equal division of the property acquired by either party during the relationship and prescribes continuing support obligations in appropriate circumstances. 54 Like married couples, New Zealand de facto couples would have the right to opt out of property division, provided that the result is not "clearly unfair" to one of the parties. 1- 7 The government's decision to treat nonmarital couples equally with married couples, rather than create a special status for them, was widely ap- 15 6 plauded in the New Zealand press.

146 Family Relations Act of British Columbia, R.S.B.C., ch. 128, § 1 (1996) (amended October 1, 1998) (Can.). 147 Property (Relationships) Act, 1984, N.S.W. Acrs § 4(1) (Austl.) (amended to include same-sex relationships by the Property (Relationships) Legislation Amend- ment Act, 1999, N.SAV A=TS § 3 (Austl.)). 148 SeeiU § 4(2). 149 See RECOGNISING SAME-SEx RELATIONSHIPS, supra note 74, at 5. 150 Id. 151 See RECOGNISING SAM]E-SEx RELATIONSHIPS, supra note 74, at 8. 152 Property (Relationships) Amendment Act, 2001, § 2D (N.Z.), available at http://wwv.brookers.co.nz/property.act/defalukhtm (last %isitedAug. 30, 2001). 153 & § 3(d). 154 Id. §§ 11, 32. 155 Id- §§ 21, 21J. 156 ,See New Property Law One Step Forward,N.Z. HERALD, Apr. 4, 2000, at A10. 1302 NOTRE DAME LAW REVIEW [VOL- 76,5

The ALI Principlesand our cultural cousins have taken the same status-based approach to nonmarital cohabitation, but their legislation is more simple and straightforward than the ALI's treatment. Even though the ALl may be understood to set a new course for American law, Chapter 6 proceeds cautiously and does not entirely divorce itself from the spirit of Marvin. Other than when the parties have had a child together, the ALl works by presumption rather than by absolute rule. In the same circumstances, Ontario and British Co- lumbia proceed by absolute rule. Similarly, facts that support a finding that the parties shared life together as a couple for purposes of the ALI formulation would equally support a loose finding of implied contract under the doctrine of Marvin.

D. Why Has the United States Been So Slow to Adopt a Status-Based Approach to Nonmarital Cohabitation? American family law has been an inspiration to our cultural cousins, leading the way in uniform child support guidelines and in marital property distribution at divorce. However, there are two puzzling anomalies of American family law. The first is our divergent treatment of inter se obligations arising from long-term stable marriage- like cohabitation, where we conduct a generally fruitless search for an agreement of the parties, rather than concentrate on the nature of the parties' de facto relationship. The second is our treatment of premarital agreements, which has become increasingly divergent over the past several decades. Our cultural cousins have long been and remain circumspect about such agreements. They do not treat the property distribution terms of premarital agreements as per se unenforceable, but they subject them to judicial review at divorce and may disregard or modify an agreement if its enforcement would work an injustice. 117

157 See the English Matrimonial Causes Act, 1973, § 24 (Eng.), which allows the divorce court to make "an order varying for the benefit of the parties to the marriage and of the children of the family or either or any of them any ante-nuptial or post- nuptial settlement... notwithstanding that there are no children of the family." Id. For discussion of this provision, see 13 HASBuRy's LAWs OF ENGLAND 1136 (4th ed. 1975). Similarly, British Columbia allows "judicial reapportionment on [the] basis of fairness" when "the provisions for division of property between spouses under their ... marriage [premarital or marital] agreement ... would be unfair" having regard to a wide variety of equitable factors. Family Relations Act of British Columbia, 1996, R.S.B.C., ch. 128, § 65 (Can). The English and the British Columbia provisions, by implication, limit enforceability to the property terms of a premarital or marital agreement. Terms relating to spousal or child support are unmentioned and, presumably, unenforceable. Com- pare the Uniform Premarital Agreement Act (UPAA), which enforces waivers of 2001] THE REGULARIZATION OF NONMARITAL COHABITATION 1303

By contrast, the Uniform Premarital Agreement Act (UPAA), adopted in some American jurisdictions, enlarges the purview of premarital agreements to include spousal support as well as property distribution and makes a premarital agreement more enforceable, in some respects, than an ordinary business contract.'5 8 It is tempting to con- clude that both anomalies simply express an American preference for contractual resolution of interpersonal issues. We are preeminently the land of free markets and individual self-determination. A purely contractual approach to nonmarital cohabitation and premarital vuri- ation of the incidents of marriage merely expresses that preference and hence forms a logical and comprehensible part of the American legal fabric. That is the story I have long told myself- our law may not be good law, but it is culturally understandable and coherent. However, a glance at our cultural cousins suggests that there may be other explanations. They too have a well-established culture of contract and self-determination. The significant difference between them and us may lie in the comparative structure and transparency of our various welfare states. In the course of writing this Essay I have spousal support, as well as property rights, without any consideration of "firness" at dissolution. UNiF. PmAtAPrrAL A REsrr Act § 6, 9a U.L.A. 369 (1987). For discussion of state law adoption and modification of these UPAA provisions, see ALI PtuxNC. PLxs (Tentative Draft 2000), supra note 24, § 7.05, reporter's n. cmn.b. For criticism of the UPAA, see i. and text immediately infra. 158 Section 6 of the UPAA provides: (a) A premarital agreement is not enforceable if the party against whom enforcement is sought proves that: (1) that party did not execute the agreement voluntarily; or (2) the agreement was unconscionable when it was executed and, before execution of the agreement, that party:. (i) was not provided a fair and reasonable disclosure of the property or financial obligations of the other party, (ii) did not voluntarily and expressly waive, in writing, any right to disclosure of the property or financial obligations of the other party beyond the disclosure provided; and (iii) did not have, or reasonably could not have had, an adequate knowledge of the property or financial obligations of tie other party. UNIF. PRxsiwuurAL AGREmrT Acr § 6, 9B U.LA 369, 376. By contrast, a business contract is unenforceable if it is merely unconscionable. To further shore up section 6, subsection (c) provides that "[a]n issue of unconscio- nability of a premarital agreement shall be decided by the court as a matter of la%,.." Id!. Of course, the UPAA represents the most extreme position with respect to the enforceability of premarital agreements. However, it reflects a trend in American law and has itself had considerable influence in the recent development of state case law. For a broad survey, see ALI PwuNciPLEs (Tentative Draft 2000), supra note 24, ch. 7. 1304 NOTRE DAME LAW REVIEW [VOL- 76:5 come to suspect that these two strains of American law may be funda- mentally ill-conceived and unjustifiable, even in terms of our own cultural values, because they are based on three interlocking fallacies. The first is that a society that seeks to limit the public provision of social welfare is also one that has limited interest in the social welfare of its citizens. (The view tends to equate the state's public assistance liability with the limits of the state's welfare concerns.) The second fallacy is the failure to recognize that we do indeed have a significant, albeit largely hidden, welfare state and to recognize the roles that various institutions play in the construction of that welfare state.16" The third, and closely related, fallacy is that at least some of us are truly self-sustaining. However, even those of us employed in the primary labor market do not fully provide for ourselves, as we like to tell ourselves. Instead, we are part of a massive government-subsidized scheme that channels welfare benefits to ourselves and our family 60 members. On one level, American family law appears to understand that the family plays a vital welfare function. Compared to our neighbors, we have an unusually highly developed law of wealth redistribution at divorce.16' However, the American discourse rationalizing that divorce

159 The usual American account is that we essentially do not have a welfare state and that everyone is required to earn his way by the "sweat of his brow." Any back-up provision, meager at that, is made only for children, the aged, and the disabled. How- ever, we do have a welfare state of significant proportions, albeit one that is largely concealed and poorly cobbled together. Itjust does not look like the welfare states of our neighbors. 160 The particular nature of the American welfare state may serve various social policies. Most obviously, it enforces a norm of full-time market employment for at least one family member, in order to secure vital welfare benefits otherwise generally unavailable to an individual or family other than through public assistance. 161 A comparative approach to the welfare function of divorce law suggests an in- verse relationship between the extent of public provision of social security and the development of an (at least apparently) robust and elaborate law of private family rights and obligations. One measure of the extent to which a country makes public provision for tie consequences of family breakdown is the degree to which it manages, by public provision, to reduce the number of children in one-parent families who would otherwise live in poverty. In first-world (OECD) countries, children in one-parent families are similarly at high risk of poverty before government transfers are taken into account. OECD standards, used for international comparison of developed countries, show a U.S. poverty rate of sixty percent for one-parent families, as compared to eleven percent for two-parent families. See Lee Rainwater & Timothy M. Smeeding, Doing Poorly: The Real Income of American Children in a Comparative Perspective 12 (Luxembourg In- come Study Working Paper, No. 127, 1995). However, by government transfers, half of OECD countries manage to reduce by more than seventy-five percent the number of children in one-parent families who would otherwise be in poverty. Id. at fig. 6. 2001] THE REGULARIZATION OF NONMARITAL COHABITATION 1305 law tends to ignore its public welfare function and instead justifies redistribution in terms of the equities between the parties. To the extent that either rationale would suffice, no harm is done by this lacuna. Nevertheless, focusing solely on the equities between the parties can be harmful when that rationale may not be fully persuasive. For example, fairness between the parties may ordinarilyjustify property distribution at divorce. However, when a party has by premarital agreement waived the right to property distribution at divorce, en-

The rest manage reductions between twenty-five and fifty percent, uith the exception of Canada, Australia, and the United States. Id. The United States reduces poverty by public transfer by the very least, by about ten percent. Id. The United States has long had the least developed public welfare state and the most elaborate and extensive law of family obligations at divorce. Canada and Austra- lia place second and third in both respects. In recent decades, Canada and Australia have adopted many aspects of American family law, particularly property division rules and uniform child support guidelines. Correspondingly, countries that are successful in avoiding poverty by public transfer appear to have less developed and elaborate private law applicable at family breakdown. This is not to suggest that private transfer rules are an effective substitute for public transfers. On the contrary, the data indicate othenise. Nevertheless, in those countries that have reduced public welfare functions in order to revitalize their econ- omies, the concurrent "privatization" of family support through the establishment of private support obligations may be read to express a public concern about adequate support and an intention to provide support by alternate, albeit private, means. To the extent that the government continues to act as a minimal provider of last resort, concern for the public fisc is also expressed in the articulation of family obligations. In countries that have reduced, or seek to reduce, their public welfare functions, this direct relationship between public welfare and private obligations is evident throughout the skein of family obligations, including property division, spousal support, and child support. For such countries, it is thus a short and inevitable step to expand the definition of "spouse" to include, first, opposite-sex couples who live in stable cohabitation and, next, same-sex couples who do likeise. In the United States, this account describes only our experience with child support. The only categorical program that contemplates income support for broken families is the child-focused Temporary Assistance for Needy Families (TANF) program, the successor to Aid to Families with Dependent Children (AFDC). Se supra note 9. The program is universal in that it covers all children ithout regard to the marital status of their parents or the social relationship, if any, of their parents. Cor- respondingly, despite prior law to the contrary, the American law of parental child support obligations has developed into a universal s)stem that makes no distinction between children born in- and out-of-wedlock. Even so, the public welfare concern most often expressed in federal AFDC and TANF legislation is to protect the fisc from avoidable public expenditure (see also the UPPA proviso for spousal support when a former spouse would othenise go on public assistance, discussed iinfra in the text at note 162), rather than the public welfare concern in assuring that adequate resources are devoted to the next generation of citizens. The absence of tie latter concern is also reflected in a history of public payments grossly insufficient to accomplish that goal. 13o6 NOTRE DAME LAW REVIEW [VOL, 76:5 forcement of that waiver is arguably "fair as between the parties" in light of their premarital agreement. On the other hand, enforcement of that contract may entirely undermine the welfare function of divorce law. Only by ignoring that welfare function can the law of premarital agreements blithely enforce the contract without further examination. In fact, the UPAA drafters might reply, they do take into account public welfare by providing that a premarital support waiver is enforceable except to the extent required to keep a divorcing spouse off public assistance.' 62 However, protection of the public treasury does not exhaust the limits of the state's interest in the social welfare of its citizens. Property division and spousal support are not limited to amounts necessary to keep a former spouse off public assistance. Nor are tax-subsidized employee health and pension benefits so limited. The fallacy here is the assumption that the welfare function of divorce-law economic redistribution extends no further than protecting the public treasury. Similarly, a law of family dissolution that exempts nonmarital cohabitants from coverage, as the Marin contractual rubric often effectively does, is insensitive to the welfare function of family law. To fulfill its welfare function, family law should cover all marriage-like relationships, the ceremonial and the nonceremonial alike. It is only when the welfare function of family law is ignored that the fact of marriage, that is, the existence of an explicit contractual undertaking, can assume decisive significance. That all stable conjugal relationships of substantial duration should be subject to the coverage of family law seems self-evident to many of our neighbors. Why is this not equally self-evident to us? In many ways we are like our neighbors. All Western countries have experienced extraordinary growth in the rate of nonmarital cohabitation during the last three decades. 163 All have responded with some form of legal regulation, although most with legal regulation that is

162 Section 6 of the UPAA provides: (b) If a provision of a premarital agreement modifies or eliminates spousal support and that modification or elimination causes one party to the agreement to be eligible for support under a program of public assistance at the time of separation or marital dissolution, a court, notwithstanding the terms of the agreement, may require the other party to provide support to the extent necessary to avoid that eligibility. UNII. PIEMAxrrAL AGREEMNT Acr § 6(b), 9B U.LA. 369, 376 (1987). 163 For American data, see supra notes 10-11. For European data, see Kiernan, supra note 119. 2001] THE REGULARIZATION OF NONMARITAL COHABITATION 1307 quite different from our own.'6 What distinguishes us is our historical and cultural experience with social welfare and, in particular, the state's role as a provider of social welfare.165 Looking abroad helps us to recognize the extent to which the American family (whether marital or nonmarital) serves vital economic functions that, in other countries, are more often assumed by society at large. In other words, a comparative perspective points up the relative importance of Ameri- can private law in assuring the welfare of children and adults at family dissolution. There is, however, a paradox. Although the family plays a greater role in the well-being of its members in the United States than it does in nations that have a more highly developed and trans- parent public welfare system, the American state's relatively weak and cloaked role as a provider of social welfare seems to obscure the welfare function of the American family. More fully developed welfare states tend to be more self-conscious about the welfare function of the family, even as they rely on it less heavily as a source of welfare for family members. Thus, when faced with growing economic pressure on their provision of welfare services (largely resulting in the decline of the welfare state), such highly developed welfare states have responded with a facially robust private law of rights and responsibilities at divorce and at the termination of nonmarital cohabitation. This pattern is evidenced in countries that are culturally closest to the United States, such as Canada, New Zealand, Australia, and England. In such countries, the trend has been to substitute private obligations for public obligations, that is, to substitute property division and family support obligations for public social security.1cG In such cases, the substitutionary role of private law ultimately paved the way for uni-

164 See text accompanying supra notes 137-57. 165 See text accompanying supra notes 5-9 and supra note 161. 166 New Zealand provides a good example. Until the early 1980s, New Zealand was a model welfare state, with a highly redistributive tax-and-benefits wealth transfer system. See Ian Shirley et al., Family Change and Family Poliies: New Zealwnd, in Fx .tLv CHANGE AND FAMILY PoUCI=s IN GREAT BRAiN, NEW ZEALND, AND THE UNrED STATES 207, 212-14 (Sheila B. Kammerman & Alfredj. Kahn eds., 1997). Plagued by a declining economy, in 1984 New Zealand undertook a radical program of privatiza- don and reduction of its tax-and-benefits wealth transfer system, often characterized as "Rogernomics" after the Labour minister, Roger Douglas, who initiated the program. Each successive amendment of the Marital Property Act has strengthened the role of property distribution and family support at dissolution of family relationships. See id. at 212-21. Similarly, "Thatcherism" in England was accompanied by the "priuatization" of child and spousal support duties. See supra notes 16-17. 1308 NOTRE DAME LAW REVIEW [VOL- 76:5 form inclusion of all cohabiting couples, whether formally married or 7 not.16 Unlike most other Western countries, the United States has never committed itself to the comprehensive goals of a fully developed welfare state. 16s Consequently, it is not ordinarily thought to be the role of the government to guarantee the social welfare of its citizenry. This perspective may have affected the way the United States has conceptualized and rationalized family law obligations, as compared to countries that have experienced the content and ethos of a more fully realized welfare state. Specifically, American family law does not recognize or acknowledge the extent to which the law of private family obligations serves a public function. This blind spot does not follow logically from the view that the government is not obliged to guarantee the social welfare of its citizenry. We could nevertheless subscribe to he notion that one reason to assign private family obligations is to serve the public interest in the economic welfare of all citizens. In other words, we could endorse the public interest goal without taking the view that the state should subvent its cost. Nevertheless, our his- toric disinclination to acknowledge that we publicly subvent that cost may prevent us from seeing that public welfare is an important goal of family law generally. Thus, our law of family support obligations generally seeks to determine what is "fair" as between the parties, albeit increasingly in a generalized rather than individualized way. Our family law literature often strains to rationalize, as "fairness between the parties" the obligations imposed by private family law. This is not to suggest that "fairness between the parties" is not an important ele- ment in the law of family obligations. But it is not a full explanation of the significance of such obligations. As a rationale for obligations, it may be incomplete but unproblematic when the issue is merely rationalizing a general statutory obligation to share the fruits of marital labor at divorce or to support a former spouse and the children of the marriage after divorce. But it is distinctly inadequate in two important areas of family law, namely (i) the enforceability of agreements (usually premarital, but occasionally marital) waiving rights otherwise ac- cruing to married persons and (ii) the treatment of stable nonmarital cohabitation. The unusual content of American law, as compared to that of our neighbors, on both subjects may be explained by our ex- clusive focus on the question: what is fair as between the parties? Dif- fering perspectives on the role of the'state and the concealed nature of our particular welfare state may explain significant differences be-

167 See text accompanying supra notes 137-63. 168 See text accompanying supra notes 5-9. 2001] THE REGULARIZATION OF NONMARITAL COHABITATION 1309

tween the United States and our cultural cousins in the legal treatment of nonmarital cohabitation and premarital agreements.

CONCLUSION This Essay has described and analyzed, at times from a comparative perspective, three significant American developments concerning nonmarital cohabitation. It has organized those developments within the rubric of rights and responsibilitiesand has attempted to link aspects of the developments to the unusual character of the American welfare state. The linkage of employment and basic welfare provision in the American employee (or shadow) welfare state 69 explains the Ameri- can initiative to obtain workplace recognition of nonmarital cohabitation as a variant of marriage. The same linkage of employment and basic welfare provision may also explain a corresponding de-linkage of family rights and responsibilities, for in the American employee welfare state, welfare rights are "earned" and welfare responsibilities are, contradistinctively, contractually undertaken (or not undertaken). The de-linkage of rights and responsibilities is most prominent in the Marvin contractual treatment of nonmarital cohabitation and the UPAA treatment of premarital agreements. Nevertheless, recent American developments may also be read to suggest that the United States is likely to join its cultural cousins in linking family rights and responsibilities. In practice, initiatives to secure employment-related benefits for nonmarital family members frequently raise the issue of corresponding responsibilities. 170 Moreover, the increasing availability of rights for nonmarital family members en- courages nonmarital cohabitants to register themselves as a family unit and should correspondingly make it difficult for them to repudi- ate that status at family dissolution by death or inter vivos separation, at least under a rubric that examines their family behavior, as opposed merely to their contractual undertakings. 171 Finally, the quest for same-sex marriage has prompted a fresh look at the rights and responsibilities of marriage by persons long denied access to that institution and has begun to generate a thoughtful literature on the meaning of

169 See text accompanying supra notes 6-8. 170 See text accompanying supra notes 89-102. 171 On the other hand, the Marvin rubric has not been accompanied by the development of a doctrine of "marriage by estoppel." Unsuccessful Maruin cases frequently reveal that the defendant filed joint (marital) tax returns uith his cohabitant and represented her as his wife for purposes of health benefits (often for childbirth). Such facts have not been held to estop the defendant from denying an agreement to be treated as a married person. See cases collected in ALI PjuNciptI.s (Tentative Draft 2000), supra note 24, § 6.03, Reporter's Note to comment b. 1310 NOTRE DAME LAW REVIEW [IVOL. 76:5 marriage and the linkage of family rights and responsibilities.172 Just as same-sex couples have spurred the movement to regularize nonmarital cohabitation, so their fresh and sustained examination of an institution that most Americans have long taken for granted may shed useful light on the social and welfare functions of the family, whether marital or nonmarital.

172 See sources cited supra note 33; see also Chambers, supra note 11, at 452-61 (Federal and state laws regarding marriage aim to "recognize affective or emotional bonds," aid in "creating an environment that is especially promising or appropriate for the raising of children," or demonstrate "assumptions (or prescriptive views) about the economic arrangements" between partners.); Donovan, supra note 85, at 652-55 (emphasizing the role of the public expectations that flow from legal recognition of relationships); Raymond C. O'Brien, Domestic Partnership:Recognition and Re- sponsibility, 32 SAN DIEGo L. REv. 163, 163 (1995) ("To date, partnerships have conferred benefits only; the most logical progression is for partnerships to include responsibilities of support, commitment and obligation within the economic partnership construct of emerging family law.").